Problems and Solutions

in
Hilbert space theory,
wavelets
and
generalized functions
by
Willi-Hans Steeb
International School for Scientific Computing
at
University of Johannesburg, South Africa

Preface
The purpose of this book is to supply a collection of problems in Hilbert
space theory, wavelets and generalized functions.
Prescribed books for problems.
1) Hilbert Spaces, Wavelets, Generalized Functions and Modern Quantum
Mechanics
by Willi-Hans Steeb
Kluwer Academic Publishers, 1998
ISBN 0-7923-5231-9
2) Classical and Quantum Computing with C++ and Java Simulations
by Yorick Hardy and Willi-Hans Steeb
Birkhauser Verlag, Boston, 2002
ISBN 376-436-610-0
3) Problems and Solutions in Quantum Computing and Quantum Information, second edition
by Willi-Hans Steeb and Yorick Hardy
World Scientific, Singapore, 2006
ISBN 981-256-916-2
http://www.worldscibooks.com/physics/6077.html
The International School for Scientific Computing (ISSC) provides certificate courses for this subject. Please contact the author if you want to do
this course or other courses of the ISSC.
e-mail addresses of the author:
[email protected]
[email protected]
Home page of the author:
http://issc.uj.ac.za

Contents
Notation

ix

1 General

2 Finite Dimensional Hilbert Spaces

3 Hilbert Space L2 ()
3.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . .
3.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . .

19
19
46

4 Hilbert Space `2 (N)

48

5 Fourier Transform

52

6 Wavelets

56

7 Linear Operators

60

8 Generalized Functions
8.1 Solved Problems . . . . . . . . . . . . . . . . . . . . . . . .
8.2 Supplementary Problems . . . . . . . . . . . . . . . . . . . .

65
65
79

Bibliography

82

Index

84

vii

Notation
:=

N
Z
Q
R
R+
C
Rn
Cn
H
i
<z
=z
|z|
T S
ST
ST
f (S)
f g
x
xT
0
k.k
x y x y
xy
A, B, C
det(A)
tr(A)
rank(A)
AT

is defined as
belongs to (a set)
does not belong to (a set)
intersection of sets
union of sets
empty set
set of natural numbers
set of integers
set of rational numbers
set of real numbers
set of nonnegative real numbers
set of complex numbers
n-dimensional Euclidean space
space of column vectors with n real components
n-dimensional complex linear space
space of column vectors with n complex components
Hilbert
space

1
real part of the complex number z
imaginary part of the complex number z
modulus of complex number z
|x + iy| = (x2 + y 2 )1/2 , x, y R
subset T of set S
the intersection of the sets S and T
the union of the sets S and T
image of set S under mapping f
composition of two mappings (f g)(x) = f (g(x))
column vector in Cn
transpose of x (row vector)
zero (column) vector
norm
scalar product (inner product) in Cn
vector product in R3
m n matrices
determinant of a square matrix A
trace of a square matrix A
rank of matrix A
transpose of matrix A
ix

A
A
A
A1
In
I
0n
AB
AB
[A, B] := AB BA
[A, B]+ := AB + BA
AB
AB
jk


t

conjugate of matrix A
conjugate transpose of matrix A
conjugate transpose of matrix A
(notation used in physics)
inverse of square matrix A (if it exists)
n n unit matrix
unit operator
n n zero matrix
matrix product of m n matrix A
and n p matrix B
Hadamard product (entry-wise product)
of m n matrices A and B
commutator for square matrices A and B
anticommutator for square matrices A and B
Kronecker product of matrices A and B
Direct sum of matrices A and B
Kronecker delta with jk = 1 for j = k
and jk = 0 for j 6= k
delta function
Heavisides function
eigenvalue
real parameter
time variable
Hamilton operator

Chapter 1

General

Problem 1.
u, v H.
(i) Show that

Let H be a Hilbert space with scalar product h , i. Let

|hu, vi| kuk kvk.

(ii) Show that
ku + vk kuk + kvk.
Problem 2. Consider a Hilbert space H with scalar product h , i. The
scalar product implies a norm via kf k2 := hf, f i, where f H.
(i) Show that
kf + gk2 + kf gk2 = 2(kf k2 + kgk2 ).
(ii) Assume that hf, gi = 0, where f, g H. Show that
kf + gk2 = kf k2 + kgk2 .

Problem 3.

Let f, g H. Use the Schwarz inequality

|hf, gi|2 hf, f ihg, gi = kf k2 kgk2

to prove the triangle inequality

kf + gk kf k + kgk.
1

2 Problems and Solutions

Problem 4.

Let H be a Hilbert space and f, g H. Show that

hf |f ihg|gi

1
(hf |gi + hg|f i)2 .
4

Problem 5. Consider a complex Hilbert space H and |1 i, |2 i H. Let

c1 , c2 C. An antilinear operator K in this Hilbert space H is characterized
by
K(c1 |1 i + c2 |2 i) = c1 K|1 i + c2 K|2 i.
A comb is an antilinear operator K with zero expectation value for all states
|i of a certain complex Hilbert space H. This means
h|K|i = h|LC|i = h|L| i = 0
for all states |i H, where L is a linear operator and C is the complex
conjugation.
(i) Consider the two-dimensional Hilbert space H = C2 . Find a unitary
2 2 matrix such that
h|U C|i = 0.
(ii) Consider the Pauli spin matrices with 0 = I2 , 1 = x , 2 = y ,
3 = z . Find
3 X
3
X
h| C|ig , h| C|i
=0 =0

where g , = diag(1, 1, 0, 1).

Problem 6. Let P be a nonzero projection operator in a Hilbert space
H. Show that kP k = 1.
Problem 7. A family, { j }jJ of vectors in the Hilbert space, H, is
called a frame if for any f H there exist two constants A > 0 and
0 < B < , such that
X
Akf k2
hj |f i|2 Bkf k2 .
jJ

Consider the Hilbert space H = R2 and the family of vectors



 

1
1
, 1 =
.
0 =
1
1
Show that we have a tight frame.

General

Problem 8. Let T : X Y be a linear map between linear spaces (vector

spaces) X, Y . The null space or kernel of the linear map T , denoted by
kerT , is the subset of X defined by
kerT := { x X : T x = 0 }.
The range of T , denoted by ranT , is the subset of Y defined by
ranT := { y Y : there exists x X such that T x = y }.
Let P be a projection operator in a Hilbert space H. Show that ranP is
closed and
H = ranP kerP
is the orthogonal direct sum of ranP and kerP .
Problem 9. Let H be an arbitrary Hilbert space with scalar product
h , i. Show that if is a bounded linear functional on the Hilbert space H,
then there is a unique vector u H such that
(x) = hu, xi

for all x H.

Problem 10. Let H be an arbitrary Hilbert space. A bounded linear operator A : H H satisfies the Fredholm alternative if one of the following
two alternatives holds:
(i) either Ax = 0, A x = 0 have only the zero solution, and the linear
equations Ax = y, A x = y have a unique solution x H for every y H;
(ii) or Ax = 0, A x = 0 have nontrivial, finite-dimensional solution spaces
of the same dimension, Ax = y has a (nonunique) solution if and only if
y u for every solution u of A u = 0, and A x = y has a (nonunique)
solution if and only if y u for every solution u of Au = 0.
Give an example of a bounded linear operator that satisfies the Fredholm
alternative.
Problem 11. Let (M, d) be a complete metric space (for example a
Hilbert space) and let f : M M be a mapping such that
d(f (m) (x), f (m) (y)) kd(x, y),

x, y M

for some m 1, where 0 k < 1 is a constant. Show that the map f has
a unique fixed point in M .

4 Problems and Solutions

Problem 12. Let H be a Hilbert space and let f : H H be a monotone
mapping such that for some constant > 0
kf (u) f (v)k ku vk

u, v H.

Show that for any w H, the equation

u + f (u) = w
has a unique solution u.
Problem 13.

Let f, g H. Find all solutions to the equations

hf, gihg, f i = i.

Problem 14.

Let f, g H. Show that

kf + gk2 + kf gk2 = 2(kf k2 + kgk2 )

where the norm is implied by the scalar product of the Hilbert space.
Problem 15.

Show that
hf, gi =

1
1
kf + gk2 kf gk2
4
4

or

1
1
i
i
kf + gk2 kf gk2 + kf + igk2 kf igk2
4
4
4
4
depending on whether we are dealing with a real and complex Hilbert space.
hf, gi =

Problem 16. Given a Hilbert space H and a Hilbert subspace G of H.

The Hilbert space projection theorem states that for every f H, there
exists a unique g G such that
(i) f g G
(ii) kf gk = inf kf hk
hG

where the space G

is defined by

G := { k H : hk|ui = 0 for all u G }.

Show that if g is the minimizer of kf hk over all h G, then it is true
that f g G .

General

Problem 17. Let { n }nZ be an orthonormal basis in a Hilbert space

H. Then any vector f H can be written as
X
f=
hf, n in .
nZ

Now suppose that { n }nZ is also a basis for H, but it is not orthonormal.
Show that if we can find a so-called dual basis { n }nZ satisfying
hn |m i = (n m)
then for any vector f H, we have
X
f=
hf |n in .
nZ

Here (n m) denotes the Kronecker delta with (n m) = 0 if n = m

and 1 otherwise.
Problem 18. Let (X1 , k k1 ) and (X2 , k k2 ) be two normed spaces. Show
that the product vector spaces X = X1 X2 is also a normed vector space
if we define
kxk := max(kx1 k1 , kx2 k2 )
with x = (x1 , x2 ).
Problem 19. Let A be a linear bounded self-adjoint operator in a Hilbert
space H. Let u, v H and C. Consider the equation
Au u = v.
(i) Show that for nonreal (i.e. it has an imaginary part) v cannot vanish
unless u vanishes.
(ii) Show that for nonreal we have
k(A I)1 vk

Problem 20.

1
kvk.
|=|

Let E be the exterior of the unit disc

{ z C : |z| > 1 }

and T the unit circle

{ z C : |z| = 1 }.

6 Problems and Solutions

Let H2 (E) be the Hardy spacee of square integrable functions on T, analytic
in the region E. The inner product for f (z), g(z) H2 (E) is defined by
Z
I
1
1
dz
f (ei ) g(ei )d =
hf, gi =
f (1/z )g(z) .
2
2i T
z
Let f (z) = z 2 and g(z) = z + 1. Find the scalar product hf, gi.
Problem 21. Let O = { u1 .u2 , . . . } be an orthonormal set in a infinite
dimensional Hilbert space. Show that if
x=

cj uj

j=1

then
kxk2 =

|cj |2 .

j=1

Problem 22. Two Cauchy sequences { xk } and { yk } are said to be

equivalent if for all  > 0, there is a k(epsilon) such that for all j k()
we have d(xj , yj ) < . One writes { xk } { yk }. Obviously, is an
equivalence relationship. Show that equivalent Cauchy sequences have the
same limit.
Problem 23. Consider the sequence { xk }, k = 1, 2, . . . in R defined
by xk = 1/k 2 for all k = 1, 2, . . .. Show that this sequence is a Cauchy
sequence.
Problem 24. Let H be a Hilbert space and S be a sub Hilbert space.
Show that any element u of H can be decomposed uniquely
u=v+w
where v is in S and w is in S .
Problem 25.

Let u, v1 , v2 be elements of a Hilbert space. Show that



v1 + v2
2
2
k2 + kv1 v2 k2 .
2ku v1 k + 2ku v2 k = k2 u
2

Problem 26.

Let P be the set of prime numbers. We define the set

S := {(p, q) : p, q P p q }.

General

Show that
d((p1 , q1 ), (p2 , q2 )) := |p1 q1 p2 q2 |
defines a metric.
Problem 27. Consider the vector space of all continuous functions defined on [a, b]. We define a metric
d(f, g) := max |f (x) g(x)|.
axb

Let a = , b = , f (x) = sin(x) and g(x) = cos(x). Find d(f, g).

Problem 28.

The n n matrices over R form a vector space. Show that

d(A, B) :=

n X
n
X

|ajk bjk |

j=1 k=1

defines a metric.
Let n 1. Consider the continuous function

0
0 t < 1/2 1/n

fn (t) = 1/2 + n2 (t 1/2) 1/2 1/n t 1/2 + 1/n

1
1/2 + 1/n t 1

Problem 29.

Show that the sequence { fn (t) } is not a Cauchy sequence for the uniform
norm, but with any of the Lp norms (1 p < ) it is a Cauchy sequence.
Problem 30. The sequence space consists of the set of all (bounded or
unbounded) sequences of complex
x = (1 , 2 , . . .)
Thus we have a vector space. Can we define a metric in this vector space
which is implied by a norm?
Problem 31.

Let f, g H. Show that

hf |f ihg|gi

1
(hf |gi + hg|f i)2 .
4

Chapter 2

Finite Dimensional
Hilbert Spaces

Problem 1.

Consider the

1
0
,
0
0

Hilbert space R4 and the vectors

1
1
1
1
1
1
, , .
1
1
0
1
0
0

(i) Show that the vectors are linearly independent.

(ii) Use the Gram-Schmidt orthogonalization process to find mutually orthogonal vectors.
Problem 2.
basis)

Consider the Hilbert space R4 . Show that the vectors (Bell

1
1 0
,
2 0
1

1
1 0

,
0
2
1

0
1 1
,
2 1
0

0
1 1

2 1
0

are linearly independent. Show that they form a orthonormal basis in the
Hilbert space R4 .
Problem 3. Consider the Hilbert space R4 . Find all pairwise orthogonal
vectors (column vectors) x1 , . . . , xp , where the entries of the column vectors
8

Finite Dimensional Hilbert Spaces

can only be +1 or 1. Calculate the matrix

p
X

xi xTi

i=1

and find the eigenvalues and eigenvectors of this matrix

Problem 4. A sequence { fn } (n N) of elements in a normed space E
is called a Cauchy sequence if, for every  > 0, there exists a number M ,
such that kfp fq k <  for p, q > M . Consider the Hilbert space R. Show
that
n
X
1
sn =
,
n1
(j 1)!
j=1
is a Cauchy sequence.
Problem 5. Two Cauchy sequences { xk } and { yk } are said to be equivalent if for all  > 0, there is a k() such that for all j k() we have
d(xj , yj ) < . One writes { xk } { yk }. Obviously, is an equivalence
relationship. Show that equivalent Cauchy sequences have the same limit.
Problem 6. Consider the sequence { xk }, k = 1, 2, . . . in R defined
by xk = 1/k 2 for all k = 1, 2, . . .. Show that this sequence is a Cauchy
sequence.
Problem 7.

Consider the Hilbert space C2 and the vectors

 


i
1
|0i =
,
|1i =
.
i
1

Normalize these vectors and then calculate the probability |h0|1i|2 .

Problem 8.

Consider the Hilbert space Rn . Let x, y Rn . Show that

kx + yk2 + kx yk2 2(kxk2 + kyk2 ).

Note that
kxk2 := hx, yi.
Problem 9. Let |0i, |1i be an orthonormal basis in the Hilbert space C2 .
Let
|i = cos(/2)|0i + ei sin(/2)|1i
where , R.

10 Problems and Solutions

(i) Find h|i.
(ii) Find the probability |h0|i|2 . Discuss |h0|i|2 as a function of .
(iii) Assume that
 
 
1
0
|0i =
,
|1i =
.
0
1
Find the 2 2 matrix |ih| and calculate the eigenvalues.
Problem 10.

Consider the Hilbert space R2 . Show that the vectors

 



1
1
1
1

,
2 1
2 1

are linearly independent. Find

 
 
1
1
1
1

,
1
2
2 1

 

1
1
1
1

,
1
2
2 1

 


1
1
1

,
1
2 1




1
1
1
1

.
2 1
2 1
1

Show that these four vectors form a basis in R4 . Consider the 4 4 matrix
Q which is constructed from the four vectors given above, i.e. the columns
of the 4 4 matrix are the four vectors. Find QT . Is Q invertible? If so
find the inverse Q1 . What is the use of the matrix Q?
Problem 11. Consider the Hilbert space R4 . Let A be a symmetric 4 4
matrix over R. Assume that the eigenvalues are given by 1 = 0, 2 = 1,
3 = 2 and 4 = 3 with the corresponding normalized eigenfunctions

1
1 0
u1 = ,
2 0
1

1
1 0
u2 =
,
0
2
1

0
1 1
u3 = ,
2 1
0

Find the matrix A by means of the spectral theorem.

Problem 12.

Show that the 2 2 matrices





1
1
1 0
0 1
A=
, B=
,
2 0 1
2 1 0




1
1
0 i
1 0
C=
, D=
1
2 i 0
2

form an orthonormal basis in the Hilbert space M2 (C).

0
1 1
u4 =
.
2 1
0

Finite Dimensional Hilbert Spaces

11

Problem 13. Consider the Hilbert space H of the 2 2 matrices over

the complex numbers with the scalar product
hA, Bi := tr(AB ),

A, B H.

Show that the rescaled Pauli matrices j :=

1
1 =
2

0
1

1
0


,

1
2 =
2

0
i

plus the rescaled 2 2 identity matrix


1
1
0 =
2 0

1 j ,
2

i
0

0
1

j = 1, 2, 3

1
3 =
2

1
0

0
1

form an orthonormal basis in the Hilbert space H.

Problem 14. Let A, B be two n n matrices over C. We introduce the
scalar product
1
tr(AB )
= tr(AB ).
hA, Bi :=
trIn
n
This provides us with a Hilbert space.
The Lie group SU (N ) is defined by the complex n n matrices U
SU (N ) := { U : U U = U U = In , det(U ) = 1 }.
The dimension is N 2 1. The Lie algebra su(N ) is defined by the n n
matrices X
su(N ) := { X : X = X , trX = 0 }.
(i) Let U SU (N ). Calculate hU, U i.
(ii) Let A be an arbitrary complex nn matrix. Let U SU (N ). Calculate
hU A, U Ai.
(iii) Consider the Lie algebra su(2). Provide a basis. The elements of the
basis should be orthogonal to each other with respect to the scalar product
given above. Calculate the commutators of these matrices.
Problem 15.

= S1 be a Hamilton operator, where

Let H

0 1 0
~
S1 :=
1 0 1
2 0 1 0

and is the frequency.

(i) Find exp(iHt/~)(0),

where (0) = (1, 1, 1)T / 3.

12 Problems and Solutions

(ii) Calculate the time evolution of

1
S3 := ~ 0
0

0
0
0

0
0
1

using the Heisenberg equation of motion. The matrices Sx , Sy , Sz are the

spin-1 matrices, where

0 i 0
~
S2 :=
i 0 i .
2 0 i
0

Problem 16.

Consider the linear operator

2 0 0
A = 0 0 1
0 1 0

in the Hilbert space R3 . Find

kAk := sup kAxk
kxk=1

using the method of the Lagrange multiplier.

Problem 17.

Consider the Hilbert space R4 . Show that the Bell basis

1
1 0
u1 = ,
2 0
1

1
1 0
u2 =
,
0
2
1

0
1 1
u3 = ,
2 1
0

0
1 1
u4 =

2 1
0

forms an orthonormal basis in this Hilbert space.

Problem 18. Consider the Hilbert space R3 . Let x R3 , where x is
considered as a column vector. Find the matrix xxT . Show that at least
one eigenvalue is equal to 0.
Problem 19.
1 =

(i) Consider the Hilbert space C4 . Show that the matrices

1
(I2 I2 + 1 1 ),
2

are projection matrices in C4 .

2 =

1
(I2 I2 1 1 )
2

Finite Dimensional Hilbert Spaces

13

(ii) Find 1 2 . Discuss.

(iii) Let e1 , e2 , e3 , e4 be the standard basis in C4 . Calculate
1 ej ,

2 ej ,

j = 1, 2, 3, 4

and show that we obtain 2 two-dimensional Hilbert spaces under these

projections.
Problem 20.

Consider the 3 3 matrix

2 0 2
A = 1 0 0.
0 0 1

(i) The matrix A can be considered as an element of the Hilbert space of

the 3 3 matrices with the scalar product hA, Bi := tr(AB T ). Find the
norm of A with respect to this Hilbert space.
(ii) On the other hand A can be considered as a linear operator in the
Hilbert space R3 . Find die norm
kAk := sup kAxk,

x R3 .

kxk=1

(iii) Find the eigenvalues of A and AAT . Compare the result with (i) and
(ii).
Problem 21. Consider the Hilbert space R3 . Find the spectrum (eigenvalues and normalized eigenvectors) of matrix

1 2 3
A = 1 2 3.
1 2 3
Find kAk := supx=1 kAxk, where k.k denotes the norm and x R3 .
Problem 22. Find the spectrum (eigenvalues and normalized eigenvectors) of the 3 3 matrix

3 3 3
A = 3 3 3.
3 3 3
Find kAk, where k.k denotes the norms
kAk1 := sup kAxk
kxk=1

14 Problems and Solutions

kAk2 :=

p
tr(AA ).

Compare the norms with the eigenvalues. Find exp(A).

Problem 23. Consider the Hilbert space M4 (C) of all 4 4 matrices over
C with the scalar product hA, Bi := tr(AB ), where A, B M4 (C). The
-matrices are given by

0 0 0 1
0 0 0 i
0 0 1 0
0 0 i 0
1 =

, 2 =
0 1 0 0
0 i 0
0
1 0 0 0
i 0 0
0
0
0
3 =
i
0

and

0
0
0
i

i 0
0 i
,
0 0
0 0

1
0
4 =
0
0

0
0
5 = 1 2 3 4 =
1
0

0
0
0
1

0
1
0
0

1
0
0
0

0
0
1
0

0
0

0
1

0
1
.
0
0

We define the 4 4 matrices

jk :=

i
[j , k ],
2

j<k

where j = 1, 2, 3, k = 2, 3, 4 and [ , ] denotes the commutator.

(i) Calculate 12 , 13 , 14 , 23 , 24 , 34 .
(ii) Do the 16 matrices
I4 , 1 , 2 , 3 , 4 , 5 , 5 1 , 5 2 , 5 3 , 5 4 , 12 , 13 , 14 , 23 , 24 , 34
form a basis in the Hilbert space M4 (C)? If so is the basis orthogonal?
Problem 24. Find the spectrum (eigenvalues and normalized eigenvectors) of matrix

1 1 1
A = 1 1 1.
1 1 1
Find kAk, where k.k denotes the norm.
Problem 25. Let A and B be two arbitrary matrices. Give the definition
of the Kronecker product. Let uj (j = 1, 2, . . . , m) be an orthonormal basis
in the Hilbert space Rm . Let vk (k = 1, 2, . . . , n) be an orthonormal basis in

Finite Dimensional Hilbert Spaces

15

the Hilbert space Rn . Show that uj vk (j = 1, 2, . . . , m), (k = 1, 2, . . . , n)

is an orthonormal basis in Rm+n .
Problem 26.

Show that the 2 2 matrices





1
1
1 0
0 1
A=
, B=
,
2 0 1
2 1 0




1
1
0 i
1 0
, D=
C=
2 i 0
2 0 1

form an orthonormal basis in the Hilbert space M 2 (C).

Problem 27. Show that the 2 2 matrices





1 0
1 1
1
A=
, B=
, C=
0 0
0 0
1

1
0


,

D=

1
1

1
1

form a basis in the Hilbert space M 2 (R). Apply the Gram-Schmidt technique to obtain an orthonormal basis.
Problem 28.

Consider the 3 3 matrices over the real numbers

2 0 2
A = 1 0 0.
0 0 1

(i) The matrix A can be considered as an element of the Hilbert space of

the 3 3 matrices over the real nunbers with the scalar product
hB, Ci := tr(BC T ).
Find the norm of A with respect to this Hilbert space.
(ii) On the other hand the matrix A can be considered as a linear operator
in the Hilbert space R3 . Find the norm
kAk := sup kAxk,

x R3 .

kxk=1

(iii) Find the eigenvalues of A and AT A. Compare the result with (i) and
(ii).
Problem 29. Consider the Hilbert space C2 . The Pauli spin matrices
x , y , z act as linear operators in this Hilbert space. Let
= ~3
H

16 Problems and Solutions

be a Hamilton operator, where

z =

1
0

0
1

and is the frequency. Calculate the time evolution (intial value problem)
of


0 1
1 =
1 0
i.e.

dx

= [1 , H](t).
dt
The matrices 1 , 2 , 3 are the Pauli matrices, where


0 i
2 =
.
i 0
i~

Problem 30.
operator

Consider the Hilbert space C4 . Consider the Hamilton

0

:= ~ 0
H
0
i

0
0
i
0

Find the time-evolution of the operator

0 0
0 0
3 =
i 0
0 i

0
i
0
0

i
0
.
0
0

i 0
0 i

0 0
0 0

using the Heisenberg equation of motion

i~

d3

= [3 , H](t).
dt

Problem 31. Let M be any n n matrix. Let x = (x1 , x2 , . . .)T . The

linear operator A is defined by
Ax = (w1 , w2 , . . .)T
where
wj :=

n
X

Mjk xk ,

k=1

wj := xj ,

j>n

j = 1, 2, . . . , n

Finite Dimensional Hilbert Spaces

17

and D(A) = `2 (N). Show that A is self-adjoint if the n n matrix M is

hermitian. Show that A is unitary if M is unitary.
Problem 32. Consider the Hilbert space Cn . Let uj , j = 1, 2, . . . , n, and
vj , j = 1, 2, . . . , n be orthonormal bases in Cn , where uj , vj are considered
as column vectors. Show that
U=

n
X

uj vj

j=1

is a unitary n n matrix.
Consider the Hilbert space R2 . Given the vectors



 

3/2
3/2
0
, u3 =
.
u1 =
, u2 =
1/2
1/2
1

Problem 33.

The three vectors u1 , u2 , u3 are at 120 degrees of each other and are
normalized, i.e. kuj k = 1 for j = 1, 2, 3. Every given two-dimensional
vector v can be written as
c1 , c2 , c3 R

v = c1 u1 + c2 u2 + c3 u3 ,

in many different ways. Given the vector v minimize

1 2
(c + c22 + c23 )
2 1
subject to the two constraints
v c1 u1 c2 u2 c3 u3 = 0.
Problem 34. Let A, H be n n hermitian matrices, where H plays the
role of the Hamilton operator. The Heisenberg equations of motion is given
by
dA(t)
i
= [H, A(t)].
dt
~
with A = A(t = 0) = A(0). Let Ej (j = 1, 2, . . . , n2 ) be an orthonormal
basis in the Hilbert space H of the n n matrices with scalar product
hX, Y i := tr(XY ),

X, Y H.

Now A(t) can be expanded using this orthonormal basis as

2

A(t) =

n
X
j=1

cj (t)Ej

18 Problems and Solutions

and H can be expanded as
2

H=

n
X

hj Ej .

j=1

Find the time evolution for the coefficients cj (t), i.e. dcj /dt, where j =
1, 2, . . . , n2 .
Problem 35. The sequence space consists of the set of all (bounded or
unbounded) sequences of complex numbers
x = (x1 , x2 , . . .)
Thus we have a vector space. Can we define a metric in this vector space
which is not implied by a norm?

Chapter 3

Hilbert Space L2()

3.1

Solved Problems

Problem 1.

A basis in the Hilbert space L2 [0, 1] is given by



B := e2ixn : n Z .

Let


f (x) =

2x 0 x < 1/2
2(1 x) 1/2 x < 1

Is f L2 [0, 1]? Find the first two expansion coefficients of the Fourier
expansion of f with respect to the basis given above.
Problem 2.
quence

(i) Consider the Hilbert space L2 [1, 1]. Consider the se

1 if 1 x 1/n
fn (x) = nx if 1/n x 1/n

+1 if 1/n x 1

where n = 1, 2, . . .. Show that { fn (x) } is a sequence in L2 [1, 1] that is a

Cauchy sequence in the norm of L2 [1, 1].
(ii) Show that fn (x) converges in the norm of L2 [1, 1] to

1 if 1 x < 0
sgn(x) =
.
+1 if 0 < x 1
19

20 Problems and Solutions

(iii) Use this sequence to show that the space C[1, 1] is a subspace of
L2 [1, 1] that is not closed.
Problem 3. Let f L2 (R). Give the definition of the Fourier transform.
Let us call the transformed function f. Is f L2 (R)? What is preserved
under the Fourier transform?
Problem 4. Consider the Hilbert space L2 [a, b], where a, b R and b > a.
Find the condition on a and b such that
hcos(x), sin(x)i = 0
where h , i denotes the scalar product in L2 [a, b].
Hint. Since b > a, we can write b = x + , where  > 0.
Problem 5. Consider the Hilbert space L2 [0, 1]. The Legendre polynomials are defined as
P0 (x) = 1,

Pn (x) =

1 dn 2
(x 1).
2n n! dxn

Show that the first first four elements are given by

P0 (x) = 1,

P1 (x) = x,

P2 (x) =

1
(3x2 1),
2

P3 (x) =

1
(5x3 3x).
2

Normalize the four elements. Show that the four elements are pairwise
orthonormal.
Problem 6. Let R be a bounded region in n-dimensional space. Consider
the eigenvalue problem
u = u,

u(q R) = 0

where R denotes the boundary of R.

(i) Show that all eigenvalues are real and positive
(ii) Show that the eigenfunctions which belong to different eigenvalues are
orthogonal.
Problem 7.

Consider the inner product space

C[a, b] = { f (x) : f is continuous on x [a, b] }

with the inner product

Z
hf, gi :=
a

f (x)g (x)dx.

Hilbert Space L2 ()

21

This implies a norm

Z
hf, f i =

f (x)f (x)dx = kf k2 .

Show that C[a, b] is incomplete. This means find a Cauchy sequence in

the space C[a, b] which converges to an element which is not in the space
C[a, b].
Problem 8.

Consider the Hilbert space L2 [, ]. Given the function

1 0<x
0
x=0
f (x) =

1 < 0

Obviously f L2 [, ]. Find the Fourier expansion of f . The orthonormal basis B is given by



1
B := k (x) = exp(ikx) k Z .
2
Find the approximation a0 0 (x) + a1 1 (x) + a1 1 (x), where a0 , a1 , a1
are the Fourier coefficients.
Problem 9. Consider the linear operator A in the Hilbert space L2 [0, 1]
defined by Af (x) := xf (x). Find the matrix elements
hPi , APj i
for i, j = 0, 1, 2, 3, where Pi are the (normalized) Legrende polynomials. Is
the matrix Aij symmetric?
Problem 10.

Consider the Hilbert space L2 [0, 2). Let

g(x) = cos(x),

f (x) = x.

Find the conditions on the coefficients of the polynomial

p(x) = a3 x3 + a2 x2 + a1 x + a0
such that
hg(x), p(x)i = 0,

hf (x), p(x)i = 0.

Solve the equations for a3 , a2 , a1 , a0 .

Problem 11.
functions

Let b > a. Consider the Hilbert space L2 ([a, b]) and the


xa
, n = 1, 2, . . .
n (x) := sin n
ba

22 Problems and Solutions

which form an orthonormal basis in L2 ([a, b]). Find
b

Z
hm (x), xn (x)i

m (x)xn (x)dx,

m, n = 0, 1, 2, . . .

Problem 12.
functions

Let b > a. Consider the Hilbert space L2 ([a, b]) and the


xa
n (x) := cos n
, n = 0, 1, 2, . . .
ba

which form an orthonormal basis in L2 ([a, b]). Find

b

Z
hm (x), xn (x)i

m (x)xn (x)dx,

m, n = 0, 1, 2, . . .

Problem 13.
functions

Let m, n N. Consider the Hilbert space L2 ([1, 1]) the

1
, n = 1, 2, . . .
1 + nx2
which are elements in this Hilbert space. Find
fn (x) =

kfn (x) fm (x)k.

Problem 14. Consider the Hilbert space L2 (R). Give the definition
and an example of an even function in L2 (R). Give the definition and an
example of an odd function in L2 (R). Show that any function f L2 (R)
can be written as a combination of an even and an odd function.
Problem 15. The Chebyshev polynomials Tn (x) of the 1-st kind are
defined for x [1, 1] and given by
Tn (x) = cos(n arccos x),

n = 0, 1, 2, . . .

The Chebyshev polynomials Un (x) of the 2-nd kind are defined for x
[1, 1] and given by
Un (x) =

sin((n + 1) arccos x)

,
1 x2

n = 0, 1, 2, . . . .

Consider the Hilbert spaces



dx

H1 = L2 [1, 1],
,
1 x2

H2 = L2

2 1 x2 dx
[1, 1],

Hilbert Space L2 ()

23

which bases are formed by the Chebyshev polynomials of the 1-st and 2-nd
type

(1)
(1)
n 1, 0 = T0 (x) = 1
n (x) = 2Tn (x),
(2)
n (x) = Un (x),

n0
(1)

(2)

Find a recursion relation for n and n .

Problem 16. Consider the Hilbert space L2 [, ]. Obviously cos(x)
L2 [, ]. Find the norm k cos(x)k. Find nontrivial functions f, g
L2 [, ] such that
hf (x), cos(x)i = 0,

hg(x), cos(x)i = 0

and
hf (x), g(x)i = 0.
Problem 17.
polynomial p

Consider the Hilbert space L2 [0, 1]. Find a non-trivial

p(x) = ax3 + bx2 + cx + d

such that
hp, 1i = 0,
Problem 18.

hp, xi = 0,

hp, x2 i = 0.

Consider the set of polynomials

{ 1, x, x2 , . . . , xn , . . . }.

Use the Gram-Schmidt procedure and the inner product

b

Z
hf, gi =

f (x)g(x)(x)dx,

(x) > 0

to obtain the first four orthogonal polynomials when

(i) a = 1, b = 1, (x) = 1 (Legrendre polynomials)
(ii) a = 1, b = 1, (x) = (1 x2 )1/2 (Chebyshev polynomials)
(iii) a = 0, b = +, (x) = ex (Laguerre polynomials)
2
(iv) a = , b = +, (x) = ex (Hermite polynomials)
Problem 19.

Consider the function

f (x) =

X
1
cos(jx).
2j
j=0

24 Problems and Solutions

Is f an element of L2 [, ]?
Problem 20. Consider the Hilbert space L2 ([0, 1]). The shifted Legrendre
polynomials , defined on the interval [0, 1], are obtained from the Legrendre polynomial by the transformation y = 2x 1. The shifted Legrendre
polynomials are given by the recurrence formula
Pj (x) =

(2j + 1)(2x 1)
j
Pj (x)
Pj1 (x)
j+1
j+1

j = 1, 2, . . .

and P0 (x) = 1, P1 (x) = 2x 1. They are elements of the Hilbert space

L2 ([0, 1]). A function u in the Hilbert space L2 ([0, 1]) can be approximated
in the form of a series with n + 1 terms
u(x) =

n
X

cj Pj (x)

j=0

where the coeffients cj R, j = 0, 1, . . . , n. Consider the Volterra integral

equation of first kind
Z x
y(t)
dt = f (x),
0tx1

0 (x t)
with 0 < < 1 and f L2 ([0, 1]). Consider the ansatz
yn (x) = a0 x +

n
X

cj Pj (x).

j=0

to find an approximate solution to the Volterra integral equation of first

kind ( = 1/2)
Z x
y(t)

dt = f (x)
xt
0
where
2
f (x) =
x(105 56x2 + 48x3 ).
105
Problem 21. The Fock space F is the Hilbert space of entire functions
with inner product given by
Z
2
1
f (z)g(z)e|z| dxdy,
z = x + iy
hf |gi :=
C
where C denotes the complex numbers. Therefore the growth of functions
in the Hilbert space F is dominated by exp(|z|2 /2). Let f, g F with
Taylor expansions
f (z) =

X
j=0

aj z j ,

g(z) =

X
j=0

bj z j .

Hilbert Space L2 ()

25

(i) Find hf |gi and kf k2 .

(ii) Consider the special that f (z) = sin(z) and g(z) = cos(z). Calculate
hf |gi.
(iii) Let
K(z, w) := ezw ,
z, w C.
Calculate hf (z)|K(z, w)i.
Problem 22. Consider the Hilbert space L2 [0, ]. Let k k be the norm
induced by the scalar product of L2 [0, ]. Find the constants a, b such that
k sin(x) (ax2 + bx)k
is a minimum.
Problem 23.

Consider the Hilbert space L2 (R). Let

fn (x) =

x
,
1 + nx2

n = 1, 2, . . . .

(i) Find kfn (x)k and

lim kfn (x)k.

(ii) Does the sequence fn (x) converge uniformely on the real line?
Problem 24.

Let n = 1, 2, . . .. We define the functions fn L2 [0, ) by


n for n x n + 1/n
fn (x) =
0
otherwise

(i) Calculate the norm kfn fm k implied by the scalar product. Does the
sequence { fn } converge in the L2 [0, ) norm?
(ii) Show that fn (x) converges pointwise in the domain [0, ) and find the
limit. Does the sequence converge pointwise uniformly?
(iii) Show that { fn } (n = 1, 2, . . .) is an orthonormal system. Is it a basis
in the Hilbert space L2 [0, )?
Problem 25.

Consider the function f L2 [0, 1]


x for 0 x 1/2
f (x) =
1 x for 1/2 x 1

A basis in the Hilbert space is given by

n
B := 1, 2 cos(nx) :

o
n = 1, 2, . . . .

26 Problems and Solutions

Find the Fourier expansion of f with respect to this basis. From this
expansion show that

X
2
1
.
=
8
(2k + 1)2
k=0

Problem 26. A particle is enclosed in a rectangular box with impenetrable walls, inside which it can move freely. The Hilbert space is
L2 ([0, a] [0, b] [0, c])
where a, b, c > 0. Find the eigenfunctions and the eigenvalues. What can
be said about the degeneracy, if any, of the eigenfunctions?
Problem 27.
function

Consider the Hilbert space L2 [0, 1]. Find a non-trivial

f (x) = ax3 + bx2 + cx + d.

such that
hf (x), xi = 0,

hf (x), x2 i = 0,

hf (x), x3 i = 0

where h , i denotes the scalar product

Problem 28. Consider the Hilbert space L2 [0, 1]. Find a non-trivial
function f such that
hf (x), xi = 0,

hf (x), x2 i = 0,

hf (x), x3 i = 0

where h , i denotes the scalar product

Problem 29.

Consider the Hilbert space L2 [0, 1] and the polynomials

1, x, x2 , x3 , x4 .

Apply the Gram-Schmidt orthogonalization process to these polynomials.

Problem 30. Consider the Hilbert space L2 (T). Let f L2 (T). Give an
example of a bounded linear functional.
Problem 31. Consider the Hilbert space L2 (R). Show that the Hilbert
space is the direct sum of the Hilbert space M of even functions and the
Hilbert space N of odd functions. Give an example of such functions in
this Hilbert space.

Hilbert Space L2 ()
Problem 32.

27

Let a > 0. Consider the Hilbert space L2 [0, a]. Let

Af (x) := xf (x)

for f L2 [0, a]. Find the norm of the operator A. We define

kAk := sup kAf k.
kf k=1

Problem 33.

Consider the Hilbert space L2 [0, 2]. Let

g(x) = cos(x),

f (x) = x.

Find the conditions on the coefficients of the polynomial

p(x) = a3 x3 + a2 x2 + a1 x + a0
such that
hg(x), p(x)i = 0,

hf (x), p(x)i = 0.

Solve the equations for a3 , a2 , a1 , a0 .

Problem 34.

Consider the function f : R R

f (x) =

1 cos(2x)
.
x

Using LHospital rules we have f (0) = 0. Is f L2 (R)?

Problem 35. Consider the Hilbert space L2 [1, 1]. The Legendre polynomials are given by
Pj (x) :=

1 dj 2
(x 1)j .
2j j! dxj

Find the scalar product

hPj (x), Pk (x)i.
Problem 36. Consider the Hilbert space H = L2 (T). This is the vector
space of 2-periodic functions. Then
1
u(x) =
2
is a constant function which is normalized, i.e. kuk = 1. Show that the
projection operator Pu defined by
Pu f := hu, f iu

28 Problems and Solutions

maps a function f to its mean. This means
Z
Pu f = hf i,
hf i =

f (x)dx.

Problem 37. Consider the Hilbert space L2 [, ] and the vector space
of continiuous real-valued functions C[, ] on the interval [, ]. Let
k > 0 and

0 if x 0
fk (x) = kx if 0 x 1/k

1 if 1/k x
The sequence of functions fk belongs to the vector space C[, ].
(i) Show that fn in the norm of the Hilbert space L2 [, ], where

0 if x 0
(x) :=
1 if 0 < x
so that the sequence { fk } is a Cauchy sequence in the Hilbert space
L2 [, ].
(ii) Show that k gk > 0 for every g C[, ]. Conclude that C[, ]
is not a Hilbert space.
Problem 38. Let be the unit disk. A Hilbert space of analytic functions
can be defined by
(
)
Z
H :=

|f (z)|2 ds <

f (z) analytic, |z| < 1 : sup

a<1

|z|=a

and the scalar product

Z
hf, gi := lim

a1

f (z)g(z)ds.
|z|=a

Let cn (n = 0, 1, 2, . . .) be the coefficients of the power-series expansion of

the analytic function f . Find the norm of f .
Problem 39. Let Cn denote the complex Euclidean space. Let z =
(z1 , . . . , zn ) Cn and w = (w1 , . . . , wn ) Cn then the scalar product
(inner product) is given by
z w := zw = zwT
where z = (z 1 , . . . , z n ). Let En denote the set of entire functions in Cn .
Let Fn denote the set of f En such that
Z
1
|f (z)|2 exp(|z|2 )dV
kf k2 := n
Cn

Hilbert Space L2 ()

29

is finite. Here dV is the volume element (Lebesgue mesure)

dV =

n
Y
j=1

dxj dyj =

n
Y

rj drj dj

j=1

with zj = rj eij . The norm follows from the scalar product of two functions
f, g Fn
Z
1
f (z)g(z) exp(|z|2 )dV.
hf, gi := n
Cn
Let
zm := z1m1 znmn
Pn
where the multiindex m is defined by m! = m1 ! mn ! and |m| = j=1 mj .
Find the scalar product
hzm , zp i.
Problem 40. Let be a complex-valued differentiable function of in
the interval [0, 2] and (0) = (2), i.e. is an element of the Hilbert
space L2 ([0, 2]). Assume that (normalization condition)
Z 2
()()d = 1.
0

Calculate

Z
d
~ 2
() ()d
=
i 0
d
where = denotes the imaginary part.
Problem 41.
The Legrendre polynomials are defined on the interval
[1, 1] and defined by the recurrence formula
Lj (x) =

2j + 1
j
xLj (x)
Lj1 (x)
j+1
j+1

j = 1, 2, . . .

and L0 (y) = 1, L1 (x) = x. They are elements of the Hilbert space

L2 ([1, 1]). Calculate the scalar product
hLj (x), Lk (x)i
for j, k = 0, 1, . . .. Discuss.
Problem 42.

Let fn : [1, 1] [1, 1] be defined by

for 1 x 0
1
fn (x) =
1 nx for 0 x 1/n

0
for 1/n x 1

30 Problems and Solutions

Show that fn L2 [1, 1]. Show that fn is a Cauchy sequence.
Problem 43. Consider the Hilbert space L2 ([1, 1]). The Chebyshev
polynomials are defined by
Tn (x) := cos(n cos1 x),

n = 0, 1, 2, . . . .

They are elements of the Hilbert space L2 ([1, 1]). We have

T0 (x) = 1,

T2 (x) = 2x2 1,

T1 (x) = x,

T3 (x) = 4x3 3x.

Calculate the scalar products

hT0 , T1 i,

hT1 , T2 i,

hT2 , T3 i.

Calculate the integrals

Z

Tm (x)Tn (x)

dx
1 x2
1
for (m, n) = (0, 1), (m, n) = (1, 2), (m, n) = (2, 3).
Problem 44.
(i) Consider the Hilbert space L2 [0, 1] with the scalar
product h, i. Let f : [0, 1] [0, 1]

2x if x [0, 1/2)
f (x) :=
2(1 x) if x [1/2, 1]
Thus f L2 [0, 1]. Calculate the moments k , k = 0, 1, 2, . . . defined by
Z 1
k := hf (x), xk i
f (x)xk dx.
0

(ii) Show that

|k |2 <

|f (x)|2 dx.

k=0

Problem 45. Let a, b R and < a < b < +. Let f be a function

in the class C 1 (i.e., the derivative df /dt exists and is continuous) on the
interval [a, b]. Thus f is also an element of the Hilbert space L2 ([a, b]).
Show that
Z
b

lim

Problem 46.

f (t) sin(t)dt = 0.
a

Consider the Lie group





2
2
SU (1, 1) =
||

||
=
1
.

(1)

Hilbert Space L2 ()

31

The elements of this Lie group act as analytic automorphism of the disk
:= { |z| < 1 }
under

z +
z zg =
z +

where (zg)h = z(gh). Let n 2. We define

Z
Hn := { f (z) analytic in , kf k2 =
|f (z)|2 (1 |z|2 )n2 dxdy < }

and

1
+ )).
f ((
z + )/(z
Un (g)f (z) :=
(z + )n

Then Hn is a Hilbert space, i.e. the analytic functions in

L2 (, (1 |z|2 )n2 dxdy)
form a closed subspace. Un is a representation, i.e.,
Un (gh) = Un (g)Un (h)
and Un (e) = I, where e is the identity element in SU (1, 1) (2 2 unit
matrix).
Show that
1
dx dy
(1 |z|2 )2
is invariant z zg.
Problem 47. Consider the problem of a particle in a one-dimensional
box. The underlying Hilbert space is L2 (a, a). Solve the Schrodinger
equation

i~
= H
t
as follows: The formal solution is given by

(t) = exp(iHt/~)(0).
The
Expand (0) with respect to the eigenfunctions of the operator H.

eigenfunctions form a basis of the Hilbert space. Then apply exp(iHt/~).

Calculate the probability
P = |h, (t)i|2
where

 q 
1
(q) = sin
a
a

32 Problems and Solutions

and

 q 
1
(q, 0) = sin
.
a
a

Problem 48.

Let f L2 (Rn ). Consider the following operators

Ty f (x) = f (x y),

translation operator

Mk f (x) = eixk f (x),

n/2

Ds f (x) = |s|

modulation operator
1

f (s

x),

s R \ {0}

dilation operator

where x k = k1 x1 + + xn kn .
(i) Find kTy f k, kMk k, kDs f k, where k k denotes the norm in L2 (Rn ).
(ii) Find the adjoint operators of these three operators.
Problem 49.

Consider the vector space

H1 (a, b) := { f (x) L2 (a, b) : f 0 (x) L2 (a, b) }

with the norm g H1 (a, b))
kgk1 :=

kgk20 + kg/xk20 .

Consider the Hilbert space L2 (, ) and f (x) = sin(x). Find the norm
kf k1 .
Problem 50.

Let f H1 (a, b). Then for a x < y b we have

Z y
f (y) = f (x) +
f 0 (s)ds.
x

(i) Show that f C[a, b].

(ii) Show that
p
|f (y) f (x)| kf k1 |y x|.
Problem 51. Consider the Hilbert space L2 [0, ). The Laguerre polynomials are defined by
dn n x
(x e ),
n = 0, 1, 2, . . . .
dxn
The first five Laguerre polynomials are given by
Ln (x) = ex

L0 (x) = 1
L1 (x) = 1 x
L2 (x) = 2 4x + x2
L3 (x) = 6 18x + 9x2 x3
L4 (x) = 24 96x + 72x2 16x3 + x4 .

Hilbert Space L2 ()

33

Show that the function

1 x/2
e
Ln (x)
n!

n (x) =

form an orthonormal system in the Hilbert space L2 [0, ).

Problem 52. Consider the Hilbert space L2 [, ]. A basis in this
Hilbert space is given by


1
B = eikx : k Z .
2
Find the Fourier expansion of
f (x) = 1.

Problem 53.

(i) Consider the functions

f (x) =

1
,
1 + x2

g(x) =

x
.
1 + x2

Obviously f, g L2 (R). Calculate the scalar product

Z
hf, gi =
f (x)g(x)dx.

(ii) Let > 0. Consider the functions

f (t) =

sin(t)
,
t

g(t) =

1 cos(t)
.
t

Obvioulsy f (0) = 1, g(0) = 0 and f, g L2 (R). Calculate the scalar

product
Z
hf, gi =
f (t)g(t)dt.

Problem 54. Consider the Hilbert space L2 [0, 1]. Let P n be the n + 1dimensional real linear space of all polynomial of maximal degree n in the
variable x, i.e.
P n = span{ 1, x, x2 , . . . , xn }.
The linear space P n can be spanned by various systems of basis functions.
An important basis is formed by the Bernstein polynomials
{B0n (x), B1n (x), . . . , Bnn (x)}

34 Problems and Solutions

of degree n with
Bin (x) := xi (1 x)ni ,

i = 0, 1, . . . , n.

The Bernstein polynomials have a unique dual basis

{D0n (x), D1n (x), . . . , Dnn (x) }
which consists of the n + 1 dual basis functions
Din (x) =

n
X

cij Bjn (x).

j=0

The dual basis functions satisfy

hDin (x), Bjn (x)i = ij .
(i) Calculate the scalar product
hBim (x), Bjn (x)i.
(ii) Find the coefficients cij .
Problem 55.
on the disc

Consider Fourier series and analytic (harmonic) functions

D := { z C : |z| 1 }.

A Fourier series can be viewed as the boundary values of a Laurent series

cn z n .

n=

Suppose we are given a function f on T. Find the harmonic extension u of

f into D. This means
u = 0

and u = f

on D = T

where := 2 /x2 + 2 /y 2 .
Problem 56.

Consider the compact abelian Lie group U (1)

U (1) = { e2i : 0 < 1 }.

The Hilbert space L2 (U (1)) is the space L2 ([0, 1]) consisting of all measureable funcrions f () with period 1 such that
Z 1
|f ()|2 d < .
0

Hilbert Space L2 ()

35

Now the set of functions

{ e2im : m Z }
form an orthonormal basis for the Hilbert space L2 ([0, 1]). Thus every
f L2 ([0, 1]) can be expressed uniquely as
Z 1
+
X
f () =
cm e2im ,
cm =
f ()e2im d.
0

m=

Calculate
Z

|f ()|2 .

Problem 57. The Hilbert space L2 (R) is the vector space of measurable functions defined almost everywhere on R such that |f |2 is integrable.
H1 (R) is the vector space of functions with first derivatives in L2 (R). Give
two examples of such a function.
Problem 58.

Consider the Hilbert space L2 [, ]. The set of functions



1 inx
e
2
nZ

is an orthonormal basis for L2 [, ]. Let

1
K(x, t) = eitx .
2
For t fixed find the Fourier expansion of this function.
Problem 59. Consider the vector space C([0, 1]) of continouos functions.
We define the triangle function

2x 0 x 1/2
(x) :=
.
2 2x 1/2 < x 1
Let 0 (x) := x and
n (x) := (2j x k)
where j = 0, 1, 2, . . ., n = 2j + k and 0 k < 2j . The functions
{ 1, 0 , 1 , . . . }
are the Schauder basis for the vector space C([0, 1]). Let f C([0, 1]).
Then

X
f (x) = a + bx +
cn n (x).
n=1

36 Problems and Solutions

(i) Find the Schauder coefficients a, b, cn .
(ii) Consider g : [0, 1] [0, 1]
g(x) = 4x(1 x).
Find the Schauder coefficients for this function.
Problem 60.
0, 1, 2, . . . be

Let s be a nonnegative integer. Let x R and hn (n =

2

hn (x) =

(1)n
dn ex
exp(x2 /2)
.
4
dxn
2n/2 n!

Thus hn for an orthonormal basis in the Hilbert space L2 (R). Consider the
sequence
s
X
1
fs (x) =
ein hn (x)
s + 1 n=0
where s = 0, 1, 2, . . .. Show that the sequence converges weakly but not
strongly to 0.
Problem 61. Let CnN be the vector space of all nN complex matrices.
Let Z CnN . Then Z Z T , where T denotes transpose. One defines a
Gaussian measure on CnN by
d(Z) :=

1
exp(tr(ZZ ))dZ
nN

where dZ denotes the Lebesgue measure on CnN . The Fock space F(CnN )
consists of all entire functions on CnN which are square integrable with
respect to the Gaussian measure d(Z). With the scalar product
Z
f, g F(CnN )
hf |gi :=
f (Z)g(Z)d(Z),
CnN

one has a Hilbert space. Show that this Hilbert space has a reproducing
kernel K. This means a continuous function K(Z, Z 0 ) : CnN CnN C
such that
Z
K(Z, Z 0 )f (Z 0 )d(Z 0 )

f (Z) =
CnN

for all Z CnN and f F(CnN ).

Problem 62.
L2 [0, )

Consider the Hilbert space L2 [0, ) and the function f

f (x) = exp(u1/4 ) sin(u1/4 ).

Hilbert Space L2 ()
Find
Z

37

f (x)xn dx,

n = 0, 1, 2, . . . .

Problem 63.
uct

Consider the Hilbert space L2 [0, 2) with the scalar prod1

hf1 , f2 i =
2

f1 (ei )f2 (ei )d.

(i) Let f1 (z) = z and f2 (z) = z 2 . Find hf1 , f2 i.

(ii) Let f1 (z) = z 2 and f2 (z) = sin(z). Find hf1 , f2 i.
Problem 64.

Consider the Hilbert space L2 (R2 ) with the basis

2

mn (x1 , x2 ) = N Hm (x1 )Hn (x2 )e(x1 +x2 )/2

where m, n = 0, 1, . . . and N is the normalization factor. Consider the
two-dimensional potential
V (x1 , x2 ) =

a 4
(x + x42 ) + cx1 x2 .
4 1

(i) Find all linear transformation T : R2 R2 such that

V (T x) = V (x).
(ii) Show that these 2 2 matrices form a group. Is the group abelian.
(iii) Find the conjugacy classes and the irreducible representations.
(iv) Consider the Hilbert space L2 (R2 ) with the orthogonal basis
2

mn (x1 , x2 ) = Hm (x1 )ex1 /2 Hn (x2 )ex2 /2

where m, n = 0, 1, 2, . . .. Find the invariant subspaces from the projection
operators of the irreducible representations.
Problem 65. Consider the Hilbert space L2 [, ]. A basis in this
Hilbert space is given by


1 ikx

B=
e
: kZ .
2
Find the Fourier expansion of
f (x) = 1.

38 Problems and Solutions

Problem 66. Consider the Hilbert space L2 [0, 1]. Let P n be the n + 1dimensional real linear space of all polynomial of maximal degree n in the
variable x, i.e.
P n = span{ 1, x, x2 , . . . , xn }.
The linear space P n can be spanned by various systems of basis functions.
An important basis is formed by the Bernstein polynomials {B0n (x), B1n (x), . . . , Bnn (x)}
of degree n with
Bin (x) := xi (1 x)ni ,

i = 0, 1, . . . , n.
(

The Bernstein polynomials have a unique dual basis {D0 x), D1n (x), . . . , Dnn (x) }
which consists of the n + 1 dual basis functions
Din (x)

n
X

cij Bjn (x).

j=0

The dual basis functions satisfy

hDin (x), Bjn (x)i = ij .
(i) Find the scalar product
hBim (x), Bjn (x)i.
(ii) Find the coefficients cij .
Problem 67. Let V be a metric vector space. A reproducing kernel
Hilbert space on V is a Hilbert space H of functions on V such that for each
x V , the point evaluation functional
x (f ) := f (x),

f H

is continouos. A reproducing kernel Hilbert space H possesses a unique

reproducing kernel K which is a function on V V characterized by the
properties that for all f H and x V , K(x, ) H and
f (x) = hf, K(x, )iH
where h, iH denotes the inner product on H. The reproducing kernel K
uniquely determines the reproducing kernel Hilbert space H. The reproducing kernel Hilbert space of a reproducing kernel K is denoted by HK .
The Paley-Wiener space is defined by
S := { f C(Rd ) L2 (Rd ) : suppf [, ]d }

Hilbert Space L2 ()

39

is a reproducing kernel Hilbert space. The Fourier transform of f L1 (Rd )

is given by
Z
1
f (x)eixk dx,
k Rd
f(k) :=
( 2)2d R2d
where x k = x1 k1 + + xd kd is the inner product in Rd . The norm on the
vector space S inherits from that in L2 (Rd ). Show that the reproducing
kernel for the Paley-Wiener space S is the sinc function
d
Y
sin((xj yj ))
sinc(x, y) :=
,
(xj yj )
j=1

x, y Rd .

Problem 68. Can one construct an orthonormal basis in the Hilbert

space L2 (R) starting from ( > 0)
e|x|/ ,

xe|x|/ ,

x2 e|x|/ ,

x3 e|x|/ , . . . .

Problem 69. Consider the Hilbert space L2 [1, 1]. Normalize the function f (x) = x in this Hilbert space.
Problem 70.

Show that (Mehlers formula)

exp((u2 +v 2 2uvz)/(1z 2 )) = (1z 2 )1/2 exp((u2 +v 2 ))

X
zn
Hn (u)Hn (v)
n!
n=0

where Hn are the Hermite polynomials.

Problem 71. Consider the Hilbert space L2 (R). Let j, k = 1, 2, . . ..
Consider the functions
fj (x) = xj e|x| ,

fk (x) = xk e|x| .

Find the scalar product

Z

(fj (x), fk (x)) :=

fj (x)fk (x)dx =

fj (x)fk (x)dx.

Discuss.
Problem 72. Consider the Hilbert space L2 (R) and the one-dimensional
Schr
odinger equation (eigenvalue equation)


d2
2 + V (x) u(x) = Eu(x)
dx

40 Problems and Solutions

where the potential V is given by
V (x) = x2 +

ax2
1 + bx2

where b > 0. Insert the ansatz

u(x) = ex

/2

v(x)

and find the differential equation for v. Discuss. Make a polynomial ansatz
for v.
Problem 73. Consider the Hilbert space L2 (R). Let g > 0. Consider
the one-dimensional Schr
odinger equation (eigenvalue equation)


x2
d2
2
u(x) = Eu(x).
2 +x +
dx
1 + gx2
Find a solution of the second order differential equation by making the
ansatz
u(x) = A(1 + gx2 ) exp(x2 /2).
Problem 74. (i) Consider the Hilbert space L2 [1/2, 1/2]. Show that
the following sets
B1 := { k (x) = exp(2ikx), k Z }

B2 := { k (x) = 2 sin(2kx), k N }
each form an orthonormal basis in this Hilbert space.
(ii) Expand the step function

1 for x [1/2, 0]
f (x) =
1 for x [0, 1/2]
with respect to the basis B1 and with respect to the basis B2 . Show that
the two expansions are equivalent. Recall that
2 sin(x) sin(y) cos(x y) cos(x + y).

Problem 75. Consider the problem of a free particle in a one-dimensional

box [a, a]. The underlying Hilbert space is L2 [a, a]. An orthonormal
basis in L2 [a, a] is given by
(+)

()

B = { uk (q), uk (q) : k N }

Hilbert Space L2 ()

41

where
(+)
uk

1
= cos
a

(k 1/2)q
a

()
uk

1
= sin
a

kq
a


.

The formal solution of the initial value problem of the Schrodinger equation
i~

= H
t

is given by

(t) = exp(iHt/~)(0).
Let

1
(q) = sin(q/a).
a

1
(q, 0) = sin(q/a),
a

Find exp(iHt/~)
and P = |h, (t)i|2 .
Problem 76. Let n be a positive integer. Consider the Hilbert space
L2 [0, n] and the function
f (x) = ex .
Find a, b R such that
kf (x) (ax2 + bx)k
is a minimum. The norm in the Hilbert space L2 [0, n] is induced by the
scalar product.
Problem 77.

Give a function f L2 ([0, )) such that

Z
Z
f (x)dx = 1,
xf (x)dx = 1.
0

Problem 78. Consider the Hilbert space L2 [0, 2]. The linear operator
Lf (x) := df (x)/dx acts on a dense subset of L2 [0, 2]. Show that this linear
operator is not bounded.
Problem 79.

Consider the Hilbert space L2 (R3 , dx) and let

S 2 = { (x1 , x2 , x3 ) : x21 + x22 + x23 = 1 }.

In spherical coordinates this Hilbert space has the decomposition

L2 (R3 , dx) = L2 (R+ , r2 dr) L2 (S 2 , sin()dd).

42 Problems and Solutions

Let I be the identity operator in the Hilbert space L2 (S 2 , sin()dd). Then
the radial momentum operator
 
1

Pr := i~
r
r r
is identified with the closure of the operator Pr I defined on D(Pr )
L2 (S 2 , sin()dd) where


1 d
+ 2
+ 2
+

rf (r) L2 (R , r dr) lim r|f (r)| = 0

D(Pr ) = f L2 (R , r dr) : f AC(R ),
r0
r dr
and for each f D(Pr )
1 d
(rf (r))
Pr f (r) = i~
r dr
where Pr is maximal symmetric in L2 (R+ , r2 dr). Show that Pr is not selfadjoint.
Problem 80.

Consider the Hilbert space L2 [, ] and the functions

f (x) = | sin(x)|

g(x) = | cos(x)|.

Find the distance

kf (x) g(x)k
in this Hilbert space.
Problem 81. Consider the Hilbert space L2 (R). Show that the spectrum
of the position operator x
is the real line denoted by R.
Problem 82.

Consider the Hilbert space L2 (R). Is

1
n (x) = p
e2in arctan(x) ,
(1 + x2 )

nZ

an orthonormal basis in L2 (R)?

Problem 83.
tions

Consider the Hilbert space L2 [0, ). Show that the funcn (x) = ex/2 Ln (x),

n = 0, 1, 2, . . .

form an orthonormal basis in L2 [0, ), where Ln are the Laguerre polynomials defined by
 
n
X
x dn n x
(1)k n k
Ln (x) =
(x
e
)
=
x .
n! dxn
k!
k
k=0

Hilbert Space L2 ()

43

Problem 84. Consider the Hilbert space L2 (R). Show that the functions
n (x) =

2n/2

2
1
Hn (x)ex /2 ,
1/4
n!()

n = 0, 1, 2, . . .

form an orthonormal basis in the Hilbert space L2 (R), where Hn are the
Hermite polynomials
2

Hn (x) = (1)n ex

Problem 85.
L2 [a, b]. Find

dn ex
,
dxn

n = 0, 1, 2, . . .

Let b > a and n = 1, 2, . . .. Consider the Hilbert space

Z

b
2

sin
a

The functions

r
n (x) =

n(x a)
ba

2
sin
ba


.

n(x a)
ba

form an orthonormal basis in the Hilbert space L2 [a, b].

Problem 86. Consider the Hilbert space L2 [0, ) and
r Z
r Z
2
2
f (x) =
g(y) cos(yx)dy,
g(x) =
f (y) cos(yx)dy.
0
0
Let g : R+ R

g(y) = ey

Find f (x).
Problem 87. Consider the Hilbert space L2 [1, 1]. An orthonormal
basis in this Hilbert space is given by


1 ikx
B= e
: |x| , k Z .
2
Consider the function f (x) = eiax in this Hilbert space, where the constant
a is real but not an integer. Apply Parsevals relation
X
1
kf k2 =
|hf, k i|2 , k (x) = eikx
2
kZ
to show that

2
1
.
=
2
2
(a k)
sin (ax)
k=

44 Problems and Solutions

Problem 88.

Consider the function f : R R

f (x) =

x
sinh(x)

with f (0) = 1. Show that

X
x
x2
.
=1+2
(1)j 2
sinh(x)
x + (j)2
j=1

Problem 89. Consider the vector space


Z
nN
nN
H(C
) := f : C
C | f holomorphic

|f (z)| d(z) <

CnN

where z = (zjk ) with zjk = xjk + iyjk (j = 1, . . . , n; k = 1, . . . , N ) and

d(z) =

1
nN

exp(tr(zz )),

dz =

n Y
N
Y

dxjk dyjk .

j=1 k=1

Show that H(CnN is a Hilbert space with respect to the inner product
Z
hf, gi =
f (z)g(z)d(z).
CnN

Problem 90.

Let > 0 and

x1
R = x2 ,
x3

x01
R0 = x02 .
x03

(i) Show that

Z

p
1
exp(|R R0 |/ X
0
=

cos(n(
))
Jn (kr)J(kr0 ) exp( k 2 + 1/2 |x3 x03 |)kdk
n
0
2
|R R |
k + 1/
0
n=0
where n = 1 for n = 0 and n = 2 for n > 0 and Jm (kr) is ordinary Bessel
function of order m.
(ii) Consider the functions

fs,k,n (R) =

k
Jn (kr)ein+isx3
2

Hilbert Space L2 ()

45

where 0k, s, and n = 0, 1, 2, . . .. Show that

Z
Z 2 Z +
dx3
fs,k,n (R)fs0 ,k0 ,n0 (R)rdr = nn0 (s s0 )(k k 0 )
0

Problem 91. Let a > 0. Consider the Hilbert space L2 ([0, a]) and the
function fn L2 ([0, a])
1
fn (x) = e2ixn/a ,
a

n = 1, 2, . . .

Find kfn (x)k and k dfndx(x) k.

Problem 92. Let a > 0. Consider the Hilbert space L2 ([0, a]) and the
linear bounded operator
T f (x) = xf (x),

f L2 ([0, a]).

Find kAk.
Problem 93. Let a > 0. Consider the Hilbert space L2 ([0, a]) and the
functions fn L2 ([0, a])
1
fn (x) = e2ixn/a ,
a
Find kfn (x)k and kdfn (x)/dxk.

n = 1, 2, . . .

46 Problems and Solutions

3.2

Supplementary Problems

Problem 94. Let 0 r < 1. Consider the Hilbert space L2 [0, 2] and
f () L2 [0, 2]. Show that
1
2

f ()d +
0

Problem 95.

1
2

f ()
0

2 X

rj f () cos(j( ))d

j=1

1 r2
d.
1 2r cos( ) + r2

Consider the Hilbert space L2 [, ]. Find the series

f () =

cn ein

n=

where [, ] and
c0 = 1, c1 = c1 = 1, c2 = c2 =

Problem 96.

1
1
, . . . , cn = cn = .
2
n

Consider the real axis R and a cover

R = j=
j= Ij ,

Ij = [j , j+1 ), j < j+1

with dj := j+1 j = |Ij |. Let fj be a window function supported in the

interval [j dj1 /2, j+1 + dj+1 /2) such that
j=
X

fj2 (x) = 1

j=

and fj2 (x) = 1 fj2 (2j+1 x) for x near j+1 . Show that the functions
2
fj (x) sin
gj,k (x) = p
2dj

(2k + 1)(x j ) ,
2dj

k = 0, 1, 2, . . .

form an orthonormal basis of the Hilbert space L2 (R) subordinate to the

partition fj .
Problem 97.

Let
S = L2,real (R, dx/(1 + |x|3 ))

Hilbert Space L2 ()

47

and q S. Consider the operator

L(q) :=

d2
+ q.
dx2

Show that the operator L(q) defines a selfadjoint operator in the Hilbert
space L2 (R, dx).
Problem 98. Let a > 0. Consider the Hilbert space L2 ([0, a] [0, a]
[0, a]). Let nj Z with j = 1, 2, 3. Show that the functions
n1 ,n2 ,n3 (x) =

1
exp(i(n1 x1 + n2 x2 + n3 x3 ))
a3/2

form an orthonormal basis in this Hilbert space.

Chapter 4

Hilbert Space `2(N)

Problem 1. Consider the Hilbert space `2 (N). Let x = (x1 , x2 , . . .)T be

an element of `2 (N). We define the linear operator A in `2 (N) as
Ax = (x2 , x3 , . . .)T
i.e. x1 is omitted and the n+1st coordinate replaces the nth for n = 1, 2, . . ..
Then for the domain we have D(A) = `2 (N). Find A y and the domain of
A , where y = (y1 , y, . . .). Is A unitary?
Problem 2. Consider the Hilbert space `2 (N) and x = (x1 , x2 , . . .)T . The
linear bounded operator A is defined by
A(x1 , x2 , x3 , . . . , x2n , x2n+1 , . . .)T = (x2 , x4 , x1 , x6 , x3 , x8 , x5 , . . . , x2n+2 , x2n1 , . . .)T .
Show that the operator A is unitary. Show that the point spectrum of A is
empty and the continuous spectrum is the entire unit circle in the -plane.
Problem 3. Consider the Hilbert space `2 (N). Suppose that S and T
are the right and left shift linear operators on this sequence space, defined
by
S(x1 , x2 , . . .) = (0, x1 , x2 , . . .),

T (x1 , x2 , x3 , . . .) = (x2 , x3 , x4 , . . .).

Show that T = S .
48

Hilbert Space `2 (N)

Problem 4.

Find the spectrum of the infinite dimensional matrix

0
1

0
A=

In other words

Problem 5.

49

1
0
1
0

0
1
0
1
..
.

0
0
1
0

0
0
0
1
..
.

...
...

....

...

1 if i = j + 1
aij = 1 if i = j 1

0 otherwise

Let Pj (j = 0, 1, 2, . . .) be the Legendre polynomials

P0 (x) = 1,

P1 (x) = x,

P2 (x) =

1
(3x2 1), . . . .
2

Calculate the infinite dimensional matrix A = (Ajk )

Z

+1

Ajk =

Pj (x)
1

dPk (x)
dx
dx

where j, k = 0, 1, . . .. Consider the matrix A as a linear operator in the

Hilbert space `2 (N0 ). Is kAk < ?
Problem 6. Let Z be the set of integers. Consider the Hilbert space
`2 (Z2 ). Let (m1 , m2 ) Z2 . Let f (m1 , m2 ) be an element of `2 (Z2 ). Consider the unitary operators
U f (m1 , m2 ) := e2im2 f (m1 +1, m2 ),

V f (m1 , m2 ) := e2im1 f (m1 , m2 +1).

They are the so-called magnetic translation operators with phase and ,
respectively. Find the spectrum of U and V . Find the commutator [U, V ].
The so-called Harper operator which is self-adjoint is defined by
:= U + U + V + V .
H
Consider the case , irrational and , rational.
Find the spectrum of H.
of a linear operator H
is defined as the
Problem 7. The spectrum (H)
set of all for which the resolvent
1
R() = (I H)

50 Problems and Solutions

is self-adjoint, the spectrum is a
does not exist. If the linear operator H
subset of the real axis. The Lebesgue decomposition theorem states that
= pp ac sing
where pp is the countable union of points (the pure point spectrum), ac
is absolutely continuous with respect to Lebesgue measure and sing is
singular with respect to Lebesgue measure, i.e. it is supported on a set of
measure zero. Consider the Hilbert space `2 (Z) and the linear operator
= I2 I2 3 1 3 I2 I2
H
where 1 is at position 0. Find the spectrum of this linear operator.
Problem 8. Let M be any n n matrix. Let x = (x1 , x2 , . . .)T . The
linear operator A is defined by
Ax = (w1 , w2 , . . .)T
where
wj =

n
X

Mjk xk ,

j = 1, 2, . . . , n

k=1

wj = xj ,

j>n

and D(A) = `2 (N). Show that A is self-adjoint if the n n matrix M is

hermitian. Show that A is unitary if M is unitary.
Problem 9. Let be the unit disk. A Hilbert space of analytic functions
can be defined by
(
)
Z
H :=

|f (z)|2 ds <

f (z) analytic |z| < 1 : sup

a<1

|z|=a

and the scalar product

Z
hf, gi := lim

a1

f (z)g(z)ds.
|z|=a

Let cn (n = 0, 1, 2, . . .) be the coefficients of the power-series expansion of

the analytic function f . Find the norm of f .
Problem 10. Let |ni be the number states (n = 0, 1, . . .). Let k =
0, 1, . . .. Define the operators
Tk :=

X
n=0

|nih2n + k|.

Hilbert Space `2 (N)

(i) Show that Tk Tk0 = kk0 I.

(ii) Show that TP

k Tk = Pk is a projection operator.

(iii) Show that k=0 Pk = I.

(iv) Is the operator

X
Tk Tk
k=0

unitary?

51

Chapter 5

Fourier Transform

Problem 1. Consider the Hilbert space L2 (R). Find the Fourier transform of the function

if
1 x 0
1
if
x0
f (x) = ex

0 otherwise

Problem 2.

(i) Find the Fourier transform for

f (x) =

exp(|x|),
2

Discuss large and small.

(ii) Calculate
Z

> 0.

f (x)dx.

Problem 3.

Find the Fourier transform of the hat function


1 |t|
for
1 < t < 1
f (t) =
0 otherwise

Problem 4. Let f L2 (R) and f L1 (R. Assume that f (x) = f (x).

Can we conclude that f(k) = f(k)?
52

Fourier Transform
Problem 5.
form of

Consider the Hilbert space L2 (R). Find the Fourier transf (x) = ea|x| ,

Problem 6.

53

a > 0.

Consider the Hilbert space L2 (R). Let a > 0. Define

 1
|x| < a
fa (x) = 2a
0 |x| > a

Calculate

Z
fa (x)dx
R

and the Fourier transform of fa . Discuss the result in dependence of a.

Problem 7.

Consider the Hilbert space L2 (R). Let


1
if
1/2 || 1

()
=
0 otherwise

and

()
= e|| ,

> 0.

(i) Calculate the inverse Fourier transform of ()

and (),
i.e.
Z
1

eit ()
(t) =
2 R
Z
1

eit ().
(t) =
2 R
(ii) Calculate the scalar product h(t)|(t)i by utilizing the identity

2h(t)|(t)i = h()|
()i.

Problem 8.
L2 (R)

Consider the Hilbert space L2 (R) and the function f


f (x) =

1 if |x| < 1
0 if |x| 1

Calculate f f and verify the convolution theorem

f[
f = ff.
Problem 9.

Let
f() =

(1 2 ) for || 1
0
for || > 1

54 Problems and Solutions

Find f (t).
Problem 10.
fa : R R

Let a > 0. Find the Fourier transform of the function

for
a x 0
x/a2 + 1/a
for
0xa
fa (x) = x/a2 + 1/a

0
otherwise

Problem 11.

Let a > 0. Find the Fourier transform of

1
fa (t) = ea|t| .
a

Discuss the cases a large and a small. Is fa L2 (R).

Problem 12. Show that the Fourier transform of the rectangular window
of size N

1
for
0nN 1
wn =
0 otherwise
is
W (ei ) =

sin(N/2) i(N 1)/2

e
.
sin(/2)

Problem 13. Consider the Hilbert space L2 (R). Let T > 0. Consider
the function in L2 (R)

A cos(t)
for
T < t < T
f (t) =
0
otherwise
where A is a positive constant. Calculate the Fourier transform.
Problem 14. Let > 0. Show that the Fourier transform of the Gaussian
function


x2
1
g (x) =
exp 2
2
2
is again a Gaussian function
g (k) = e
We have

g (x)dx

2 2

k /2

= 1. Is
Z
gk (k)dk = 1 ?

Fourier Transform
Problem 15.

55

Show that the analytic function f : R R

f (x) = sech(x)

is an element of L2 (R) and L1 (R). Find the Fourier transform of the function.
Problem 16.

Let a > 0. Find the Fourier transform of

2fa (x) +

sin(ax)
ax

where fa is the function with 1 for |x| a and 0 otherwise.

Problem 17.

Consider the Hermite-Gauss functions

21/4
Hn ( 2x) exp(x2 ),
fn (x) =
2n n!

n = 0, 1, 2, . . .

where Hn is the nth Hermite polynomial. They for an orthonormal basis

in the Hilbert space L2 (R). Do the Fourier transform of the functions form
an orthonormal basis in the Hilbert space L2 (R).
Problem 18. The Hilbert transform h(t) of the function f (t) is the
principal value of the convolution of f (t) with the kernel function k(t) =
1/(t)
Z
Z
1
h(t) =
f (s)k(t s)ds =
f (s)
ds.
t

Let

G() =

g(t) exp(it)dt

be the Fourier transform of g. Show that the Hilbert transform can be

written as
H() = F ()K() = isgn()F ().
Problem 19.

Let 0 > 0 be a fixed frequency and t the time. Calculate

Z
1

f () =
e|0 t| eit dt.
2

Chapter 6

Wavelets

Problem 1. Consider the Hilbert space L2 [0, 1] and the function f (x) =
x2 in this Hilbert space. Project the function f onto the subspace of L2 [0, 1]
spanned by the functions (x), (x), (2x), (2x 1), where

1
for
0x<1
(x) :=
0 otherwise

for
0 x < 1/2
1
for
1/2 x < 1 .
(x) := 1

0 otherwise
This is related to the Haar wavelet expansion of f . The function is called
the father wavelet and is called the mother wavelet.
Problem 2.

Consider the function H L2 (R)

1 0 x 1/2
H(x) = 1 1/2 x 1

0 otherwise

Let
Hmn (x) := 2m/2 H(2m x n)
where m, n Z. Draw a picture of H11 , H21 , H12 , H22 . Show that
hHmn (x), Hkl (x)i = mk nl ,

k, l Z

where h . i denotes the scalar product in L2 (R) Expand the function

f (x) = exp(|x|)
56

Wavelets

57

with respect to Hmn . The functions Hmn form an orthonormal basis in

L2 (R).
Problem 3. Consider the Hilbert space L2 [0, 1] and the Haar scaling
function (father wavelet)

1
if
0x<1
(x) =
0 otherwise
Let n be a positive integer. We define

gk (x) := n(nx k),

k = 0, 1, . . . , n 1.

(i) Show that the set of functions { g0 , g1 , . . . , gn1 } is an orthonormal set

in the Hilbert space L2 [0, 1].
(ii) Let f be a continuous function on the unit interval [0, 1]. Thus f
L2 [0, 1]. Form the projection fn on the subspace Sn of the Hilbert space
L2 [0, 1] spanned by { g0 , g1 , . . . gn1 }, i.e.
fn =

n1
X

hf, gk igk .

k=0

Show that fn (x) f (x) pointwise in x as n .

Problem 4.

The continuous wavelet transform



Z
1 +
tb
W f (a, b) =
dt,
(a, b R, a > 0)
f (t)
a
a

decomposes the function f L2 (R) hierarchically in terms of elementary

components ((tb)/a). They are obtained from a single analyzing wavelet
applying dilations and translations. Here denotes the complex conjugate of and a is the scale and b the shift parameter. The function has
to be chosen so that it is well localized both in physical and Fourier space.
The signal f (t) can be uniquely recovered by the inverse wavelet transform


Z + Z +
1
t b da
f (t) =
W f (a, b)
db
C 0
a
a

if (t) (respectively its Fourier transform ()

satisfies the admissibility
condition
Z +
|()|2
C =
d < .

0
Consider the analytic function
2

(t) = tet

/2

58 Problems and Solutions

Does satisfies the admissibility condition?
Problem 5.

Consider the function H L2 (R)

1 0 x < 1/2
H(x) = 1 1/2 x 1

0 otherwise

Let
Hmn (x) := 2m/2 H(2m x n)
where m, n Z. Draw a picture of H11 , H21 , H12 , H22 . Show that
hHmn (x), Hkl (x)i = mk nl ,

k, l Z

where h . i denotes the scalar product in L2 (R). Expand the function

f (x) = exp(|x|)
with respect to Hmn . The functions Hmn form an orthonormal basis in
L2 (R).
Problem 6. Consider the Hilbert space L2 (R). Let L2 (R) and
assume that satisfies
Z
(t)(t k)dt = 0,k
R

i.e. the integral equals 1 for k = 0 and vanishes for k = 1, 2, . . .. Show that
for any fixed integer j the functions
jk (t) := 2j/2 (2j t k),

k = 0, 1, 2, . . .

form an orthonormal set.

Problem 7.

Consider the function : R R


1
for
x [0, 1]
(x) :=
0 otherwise

Find (x) := (2x) (2x 1). Calculate

Z
(x)dx.

Problem 8. Consider the Littlewood-Paley orthonormal basis of wavelets.

The mother wavelet of this set is
1
(sin(2x) sin(x)).
L(x) :=
x

Wavelets
Then
Lmn (x) =

1
2m/2

L(2m wx n),

59

m, n Z

generates an orthonormal basis in the Hilbert space L2 (R). Apply the rule
of LHospital to find L(0).
Problem 9. (i) Consider the Hilbert space L2 (R) and L2 (R). The
basic scaling function (father wavelet) satisfies a scaling relation of the form
(x) =

N
1
X

ak (2x k).

k=0

Show that the Hilbert transform of

1
H()(y) =

Z
R

(x)
dx
xy

is a solution of the same scaling relation. Note that the scaling function
may have compact support, the Hilbert transform has support on the real
line and decays as y 1 .
(ii) Show that the Hilbert transform of the related mother wavelet is also
noncompact and decays like y p1 where
Z
xm (x)dx = 0
R

for m = 0, 1, . . . , p 1.

Chapter 7

Linear Operators

Problem 1.
operator.

Show that an isometric operator need not be a unitary

Problem 2. Consider the Hilbert space L2 [0, 1]. Show that the linear
operator T : L2 [0, 1] L2 [0, 1] defined by
T f (x) = xf (x)
is a bounded self-adjoint linear operator without eigenvalues.
Problem 3. Show that if two bounded self-adjoint linear operators S and
T on a Hilbert space H are positive semi-definite and commute (ST = T S),
then their product ST is positive semi-definite. We have to show that
hST f, f i 0 for all f H.
Problem 4. Let a > 0. Consider the Hilbert space L2 [a, a]. Consider
the Hamilton operator
2
2
= ~ d + V (x)
H
2m dx2

where

V (x) =

0
for
|x| a
otherwise
60

Linear Operators

61

Solve the Schr

odinger equation, where the initial function (t = 0) = (x)
is given by

x/a2 + 1/a for a x 0
(x) =
x/a2 + 1/a for 0 x a
Normalize . Calculate the probability to find the particle in the state
 x 
1
(x) = sin
a
a
after time t. A basis in the Hilbert space L2 [a, a] is given by




 nx  1
(n 1/2)x
1
sin
, cos
n = 1, 2, . . . .
a
a
a
a

Problem 5. Show that in one dimensional problems the energy spectrum

of the bound state is always non-degenerate. Hint. Suppose that the oppsite
is true. Let u1 and u2 be two linearly independent eigenfunctions with the
same energy eigenvalues E, i.e.
d2 u1
2m
+ 2 (E V )u1 = 0,
dx2
~

d2 u2
2m
+ 2 (E V )u2 = 0.
dx2
~

Problem 6. A particle is enclosed in a rectangular box with imprenetrable walls, inside which it can move freely. The Hilbert space is L2 ([0, a]
[0, b] [0, c]). Find the eigenfunctions and eigenvalues. What can be said
about the degeneracy, if any, of the eigenfunctions.
Problem 7. Conside the Hilbert space L2 [0, 1] and the linear operator
T : L2 [0, 1] L2 [0, 1] defined by
(T f )(x) := xf (x).
Show that T is self-adjoint and positive definite. Find its positive square
root.
Problem 8.
defined by

Consider the Hilbert space `2 (N) and the linear operator T

T : (x1 , x2 , x3 , . . .) 7 (0, 0, x3 , x4 , . . .).

Is T bounded? Is T self-adjoint? If so is T positive?

Problem 9.

In classical mechanics we have

L = r p,

T=rF

62 Problems and Solutions

where T is the torque, F = V (V potential depending only on r) and
dL
= T.
dt
In quantum mechanics with p i~, r r and wave function we
have
Z
L = i~
d3 x (r )
R3

and

d3 x (r V )

T=
R3

since F = V . and obey the Schrodinger equation

i~
i~
Show that

~2 2

=
+V
t
2m

~2 2
=
+ V .
t
2m
dL
= T.
dt

be a bounded self-adjoint Hamilton operator with

Problem 10. Let H
normalized eigenfunctions j (j I) which form an orthonormal basis in
the underlying Hilbert space. We can write
X
(t) =
cj eiEj t/~ j
jI

Find P (t) = h(t = 0)|(t)i.

where Ej are the eigenvalues of H.
Problem 11.

Consider the Hamilton operator

2
2
= ~ d + D(1 ex )2 + eEx cos(t)
H
2m dx2

where > 0. Find the quantum Liouville equation for this Hamilton
operator.
Problem 12.

Consider the Schrodinger equation



1

=
+ V (x)
i~
t
2m

Linear Operators

63

Find the coupled system of partial differential equations for




:= ,
v := =
.

Problem 13. Consider the Hilbert space L2 (R). Let f L2 (R) and
R. We define the operator U () as
U ()f (x) := ei/2 f (xei ).
Is the operator U () unitary?
Problem 14. Let a > 0. Consider the Hilbert space L2 ([0, a]) and the
linear bounded operator A defined by
Af (x) := xf (x),

f L2 ([0, a])

Find kAk.
Problem 15. Consider the Hilbert space L2 (R). Let k Z. For k = 0
we define s0 = 0, for k 1 we define
sk := 1 +

1
1 1
+ + +
2 3
k

and for k < 0 we define sk = sk . Let  > 0. Define the indicator

functions Wk as

1
for
sk < x/ sk+1
Wk (x) :=
0 otherwise
Let u L2 (R). Define the linear operator O as
(Ou)(x) := g(x)u(x)
where

x X
g(x) = +


kZ

sk + sk+1
2


Wk (x).

(i) Show that O is a bounded self-adjoint operator for any  > 0.

(ii) Show that the norm of O
kOk = sup kOuk
kuk=1

is given by 1/2.

64 Problems and Solutions

Problem 16.

Consider the momentum operator

p = i

d
dx

defined on C0 [0, ). Show that the operator has no self-adjoint extensions

on the Hilbert space L2 [0, ).
Problem 17.

Consider the Hilbert space L2 (R) and the linear operator

p = i

d
dx

on D = {f : f,

df
L2 (R)}.
dx

Show that the spectrum is the whole real axis.

Chapter 8

Generalized Functions

8.1

Solved Problems

Problem 1.

Consider the function H : R R

1 0 x 1/2
H(x) := 1 1/2 x 1

0 otherwise

Find the derivative of H in the sense of generalized functions. Obviously

H can be considered as a regular functional
Z
H(x)(x)dx.
R

Find the Fourier transform of H. Draw a picture of the Fourier transform.

Problem 2. Let C m [a, b] be the vector space of m-times differentiable
functions and the m-th derivative is continuous over the interval [a, b] (b >
a). We define an inner product (scalar product) of such two functions f
and g as

Z b
dm f dm g
df dg
+ + m m dx.
hf, gim :=
fg +
dx dx
dx dx
a
Given (Legendre polynomials)
f (x) =

1
(3x2 1),
2

g(x) =
65

1
(5x3 3x)
2

66 Problems and Solutions

and the interval [1, 1], i.e. a = 1 and b = 1. Show that f and g are
orthogonal with respect to the inner product hf, gi0 . Are they orthogonal
with respect to hf, gi1 ?
Problem 3.

Let P be the parity operator, i.e.

P r := r.

Obviously, P = P 1 . We define
OP u(r) := u(P 1 r) u(r).
The vector r can be expressed in spherical coordinates as
r = r(sin cos , sin sin , cos )
where
0 < 2 0 < .
(i) Calculate P (r, , ).
(ii) Let
(1)l+m
Ylm (, ) =
2l l!

2l + 1 (l m)!
4 (l + m)!

1/2

(sin )m

dl+m
(sin )2l eim
d(cos )l+m

be the spherical harmonics. Find

OP Ylm .
Problem 4. In the Hilbert space H = `2 (N0 ) Bose annihilation and
creation operators denoted by b and b are defined as follows: They have a
common domain

X
D(b) = D(b ) = = (x0 , x1 , x2 , . . .)T :
j|xj |2 < .

j=0

Then b is given by

b(x0 , x1 , x2 , . . .)T = (x1 , 2x2 , 3x3 , . . .)T

and b is given by

b (x0 , x1 , x2 , . . .) = (0, x0 , 2x1 , 3x2 , . . .).

The infinite dimensional vectors
un = (0, 0, . . . , 0, 1, 0, . . .)T

Generalized Functions

67

where the 1 is at the n position (n = 0, 1, 2, . . .) form the standard basis in

H = `2 (N0 ). Is
= (1, 1/2, 1/3, . . . , 1/n, . . .)
an element of D(a)?
Problem 5. Given a function (signal) f (t) = f (t1 , t2 , . . . , tn ) L2 (Rn )
of n real variables t = (t1 , t2 , . . . , tn ). We define the symplectic tomogram
associated with the square integrable function f



Z
 2
n 
n
X
Y

i
iX
1 
j 2
j
dt1 dt2 dtn f (t) exp
tj
tj 
w(X, , ) =

2|k |  Rn
2j
j

j=1
k=1
where (j 6= 0 for j = 1, 2, . . . , n)
X = (X1 , X2 , . . . , Xn ),

= (1 , 2 , . . . , n ),

(i) Prove the equality

Z

Z
w(X, , )dX =

Rn

= (1 , 2 , . . . , n ).

|f (t)|2 dt

(1)

Rn

for the special case n = 1. The tomogram is the probabilty distribution

function of the random variable X. This probability distribution function
depends on 2n extra real parameters and .
(ii) The map of the function f (t) onto the tomogram w(X, , ) is invertible. The square integrable function f (t) can be associated to the density
matrix
f (t, t0 ) = f (t)f (t0 ).
This density matrix can be mapped onto the Ville-Wigner function
Z

u
u  ipu
W (q, p) =
f q + , q
e
du.
2
2
Rn
Show that this map is invertible.
(iii) How is the tomogram w(X, , ) related to the Ville-Wigner function?
(iv) Show that the Ville-Wigner function can be reconstructed from the
function w(X, , ).
(v) Show that the density matrix f (t)f (t0 ) can be found from w(X, , ).
Problem 6. Starting with the set of polynomials { 1, x, x2 , . . . , xn , . . . }
use the Gram-Schmidt procedure the scalar product (inner product)
Z 1
hf, gi =
(f (x)g(x) + f 0 (x)g 0 (x))dx
1

68 Problems and Solutions

to find the first five orthogonal polynomials, where f 0 denotes derivative.
Problem 7. Describe the one-dimensional scattering of a particle incident
on a Dirac delta function, i.e.
U (q) = U0 (q)
where u0 > 0. Find the transmission and reflection coefficient.
Problem 8. (i) Give the definition of the current density, transmission
coefficient, and reflection coefficient.
(ii) Calculate the transmission and the reflection coefficients of a particle
having total energy E, at the potential barrier given by
V (x) = a(x),

Problem 9.

a>0

Show that

X
1 X ikx
(x 2k)
e
=
2
k=

k=

in the sense of generalized functions

Hint. Expand the 2 periodic function
f (x) =

x
1

2 2

into a Fourier series.

Problem 10. (i) Give the definition of a generalized function.
(ii) Calculate the first and second derivative in the sense of generalized
function of

0
x<0

f (x) = 4x(1 x) 0 x 1

0
x>1
(iii) Calculate the Fourier transform of f (x) = 1 in the sense of generalized
functions.
Problem 11.

Consider the generalized function

f (x) = | cos(x)|.

Find the derivative in the sense of generalized functions.

Generalized Functions
Problem 12.

69

Consider the generalized function


cos(x)
for
x [0, 2)
f (x) :=
0 otherwise

Find the first and second derivative of f in the sense of generalized functions.
Problem 13.

Find the derivative of f : R R

f (x) = |x|

in the sense of generalized functions.

Problem 14.
defined by

Find the first three derivatives of the function f : R R

f (x) = e|x|

in the sense of generalized functions.

Problem 15. The Sobolev space of order m, denoted by H m (), is defined
to be the space consisting of those functions in the Hilbert space L2 () that,
together with all their weak partial derivatives up to and including those
of order m, belong to the Hilbert space L2 (), i.e.
H m () := { u : D u L2 () for all such that || m }.
We consider real-valued functions only, and make H m () an inner product
space by introducing the Sobolev inner product h, iH m defined by
Z X
hu, viH m :=
(D u)(D v)dx for u, v H m ().
||m

This inner product generates the Sobolev norm k kH m defined by

Z X
kuk2H m = hu|uiH m =
(D u)2 dx.
||m

Thus H 0 () = L2 (). We can write

X
hu, vi =
hD , D viL2 () .
||m

In other words the Sobolev inner product hu, viH m () is equal to the sum
of the L2 () inner products of D u and D v over all such that || m.

70 Problems and Solutions

(i) Consider the domain = (0, 2) and the function

x2
0<x1
u(x) =
2
2x 2x + 1 1 < x < 2.
Obviously u L2 (). Find the Sobolev space to which u belongs.
(ii) Find the norm of u.
Problem 16.

Let c > 0. Consider the Schrodinger equation

~2 d2
+ c (n) (x) = E
2m dx2

where (n) (n = 0, 1, 2, . . .) denotes the n-th derivative of the delta function.

Derive the joining conditions on the wave function .
Problem 17. The Morlet wavelet consists of a plane wave modulated by
a Gaussian, i.e.
2
1
() = 1/4 ei e /2

where is the dimensionless frequency. Show that if = 6 the admissibility

condition is satisfied.
Problem 18.

Let
f0 (x) = exp(x2 /2).

We define the mother wavelets fn as

fn (x) =

d
fn1 (x),
dx

n = 1, 2, . . .

Show that the family of fn s obey the Hermite recursion relation

fn (x) = xfn1 (x) (n 1)fn2 (x),

Problem 19.

n = 2, 3, . . . .

Let a > 0. Show that

X
m=

exp(i2m(x + q)/a) a

(x + q ka).

k=

Problem 20. Let f : R R be a differentiable function and x0 a root

of f , i.e. f (x0 ) = 0. Show that


1
f 00 (x0 )
0
0 (f (x)) = 0

(x

x
)
+
(x

x
)
0
0
(f (x0 ))2
f 0 (x0 )

Generalized Functions
Problem 21.

71

Show that the sum

1X
(2` + 1)P` (x)P` (y)
2
`=0

of Legendre polynomials P` is given by the Dirac delta function (y x)

for 1 x +1 and 1 y +1.
Problem 22.

Show that

(x x0 ) =

1X
1
+
(cos(nx) cos(nx0 ) + sin(nx) sin(nx0 )).
2 n=1

72 Problems and Solutions

Miscellaneous
Problem 23.

Let

= H
t
be the Schr
odinger equation, where
i~

2
= ~ + U (r),
H
2m

:=

2
2
2
+
+
x21
x22
x23

and r = (x1 , x2 , x3 ). Let

t)(r, t)
(r, t) := (r,
Find j such that
divj +

= 0.
t

Problem 24. A particle is enclosed in a rectangular box with impenetrable walls, inside which in can move freely. The Hilbert space is
L2 ([0, a] [0, b] [0, c])
where a, b, c > 0. Find the eigenfunctions and eigenvalues. What can be
said about the degeneracy, if any, of the eigenfunctions?
Problem 25. Show that in one-dimensional problems the energy spectrum of the bound states is always non-degenerate. Hint. Suppose that
the opposite is true. Let u1 , u2 be two linearly independent eigenfunctions
with the same energy eigenvalue E, i.e.
d2 u1
2m
+ 2 (E V )u1 = 0
2
dx
~
d2 u2
2m
+ 2 (E V )u2 = 0.
dx2
~
Problem 26.

Derive the Heisenberg uncertainty relation.

Problem 27. Give the standard postulates in quantum mechanics and

discuss the problematic.
Problem 28. Show that in one-dimensional problems the energy spectrum of the bound states is always non-degenerate.

Generalized Functions

73

Hint. Suppose that the opposite is true.

Let u1 and u2 be two linearly independent eigenfunctions with the same
energy eigenvalues E.
2m
d2 u1
+ 2 (E V )u1 = 0
2
dx
~
d2 u2
2m
+ 2 (E V )u2 = 0
dx2
~
Problem 29.

Let a > 0 and let fa : R R be given by


x/a2 + 1/a for a x 0
fa (x) =
x/a2 + 1/a for 0 x a

The function fa generates regular functional. Find the derivative of fa in

the sense of generalized functions.
Problem 30. Consider a one-dimensional lattice (chain) with lattice
constant a. Let k be the sum over the first Brioullin zone we have
Z /a
1 X
a
F ((k))
F ((k))dk = G
N
2 /a
k1.BZ

where
(k) = 0 21 cos(ka).
Using the identity
Z

(E (k))F (E)dE F ((k))

we can write
a
G=
2

(E (k))dk dE.

F (E)

/a

/a

Calculate
Z

/a

(E (k))dk

g(E) =
/a

where g(E) is called the density of states.

Problem 31.

Let  > 0. Consider the Schrodinger eigenvalue equation



d2
2 + 2(x) u(x, ) = E()u(x, )
dx

74 Problems and Solutions

with the boundary conditions u(1, ) = 0. Here  is the coupling constant and determines the penetrability of the potential barrier. Find the
eigenfunctions and the eigenvalues.
Problem 32.

Show that in the sense of generalized functions

(x) =

1
1
lim e|x|/
2 0 

sin2 (x)
1
lim 
 (x)2


1
1
|x| |x|/
(x) = lim
1+
e
.
4 0 

(x) =

Problem 33.
functions

Give two interpretations of the series of derivatives of

f (k) = 2

cn (1)n (n) (k).

(1)

n=0

Problem 34.

Show that
Z

H(x a) =

Problem 35.

du

d
exp(iu( x)).
2

Let a > 0. Show that

1
(x2 a2 ) =
((x a) + (x + a)).
2a

Problem 36. (i) Show that the Fourier transform in the sense of generalized function of the Dirac comb
X
(x n)
nZ

is again a Dirac comb.

(ii) Find the Fourier transform in the sense of generalized functions of

1 + 2(x).

Problem 37.

(i) Consider the nonlinear differential equation

3u

du
du d2 u
d3 u
=2
+
u
.
dx
dx dx2
dx3

Generalized Functions

75

Show that u(x) = e|x| is a solution in the sense of generalized function.

(ii) Consider the nonlinear partial differential equation
u
3u
3u
u
u 2 u
+
u
.

+
3u
=
2
t
x2 t
x
x x2
x3
Show that u(x, t) = c exp(|x ct|) (peakon) is a solution in the sense of
generalized functions.
Problem 38. Let f be a differentiable function with a simple zero at x =
a such that f (x = a) = 0 and df (x = a)/dx 6= 0. Let g be a differentiable
function with a simple zero at x = b 6= a such that g(x = b) = 0 and
dg(x = b)/dx 6= 0. Show that
(f (x)g(x)) =

1
1
(x a) +
(x b)
|f 0 (a)g(a)|
|f (b)g 0 (b)|

where 0 denotes differentiation.

Problem 39. Consider the non-relativistic hydrogen atom, where a0 is
the Bohr radius and a = a0 /Z. The Schrodinger-Coulomb Green function
G(r1 , r2 ; E) corresponding to the energy variable E is the solution of the
partial differential equation


~2
~2 2
1
E G(r1 , r2 ; E) = (r1 r2 )

2m
amr1
with the appropiate boundary conditions. Show that expanding G in terms
of spherical harmonics Y`m
G(r1 , r2 ; E) =

`
X
X

g` (r1 , r2 ; E)Y`m (1 , 1 )Y`m

(2 , 2 )

`=0 m=`

we find for the radial part g` of the Schrodinger-Coulomb Green function





1 d
`(` + 1)
2
1
2m (r1 r2 )
2 d
r

g` (r1 , r2 ; ) = 2
1
2
2
2
2
r1 dr1
dr1
r1
ar1
a
~
r1 r2
where 2 a2 := ~2 /(2mE).
Hint. Utilize the identity
(r1 r2 ) =

`
(r1 r2 ) X X

Y`m (1 , 1 )Y`m
(2 , 2 )
r1 r2
`=0 m=`

Problem 40.

Show that (distributional identity on L1 (R))

Z
1
1
dx = (t s)
2 R (t x)(s x)

76 Problems and Solutions

where the integral is evaluated in the principal value sense.
Problem 41. Let c > 0. Show that an integral representation of the
delta function is given by
Z c+i
1
(x) =
etx dt
2i ci
where the path of the t-integration can be closed to the right or left.
Problem 42.

Show that
Z

x
1
t
=
eit(tx/c) d.
c
2

Problem 43.

Show that in the sense of generalized functions

Z +
Z
1
exp(ikx)dx,
(x) =
(x) =
( x)d.
2
0

Problem 44.

Let  > 0. Show that



(x a)2
1
f (x a) = exp



tends to (x a) in the sense of generalized function if  0+ .

Problem 45.

Let J0 be the Bessel functions. Show that

Z
p
1
(x)(x) =
kJ0 (k (x2 + y 2 ))dk
2 0

in the sense of generalized functions.

Problem 46.

Problem 47.

Let [0, 1). Show that



Z
1

1
x
P
= cot(/2).
2
1

x
2
0

Show that

3/2


2

(x x0 ) = lim
exp
(x x0 )2 .

2( )

Generalized Functions
Problem 48.

77

Let p [0, 1] and

(x) =

1 |x|
pe
+ (1 p)(x).
2

Then (x) 0. Show that in the sense of generalized function

Z
(x)dx = 1.
R

Problem 49.

The two-dimensional Dirac comb function is defined by

XX
C(x1 , x1 ) :=
(x1 m)(x2 n).
mZ nZ

Find the Fourier transform of C in the sense of generalized functions.

Problem 50.

Let T > 0. Consider the sequence of functions

 n
1 n nt
exp(nt/T )
fn (t) =
n! T T

where n = 1, 2, . . .. Find fn (t) for n in the sense of generalized

functions. Find the Laplace transform of fn (t).
Problem 51. What charge distribution (r) does the spherical symmetric
potential
er
V (r) =
r
give? For r 6= 0 Poissons equation in spherical coordinates is given by
V (r) =

1 d2
(rV (r)) + R(, )V (r) = 4(r)
r dr2

where R(, ) is the differential operator depending on the angles , .

Let a1 , a2 , . . . , an 6= 0. Show that
Pn
Z 1
Z 1
(1 j=1 j
1
.
= (n 1)!
d1
dn Pn
a1 a2 an
( j=1 j aj )n
0
0

Problem 52.

Problem 53.
f (y)

Prove the identity in the sense of generalized function

df (y)
(x y) f (x) (x y) +
(x y).
y
x
dy

78 Problems and Solutions

Let S(R). Show that in the sense of generalized

Problem 54.
functions

lim

Problem 55.

sin(x)
(x)dx = (0).
x

Show that
u(x, t) =



1
(x x0 )2
exp
4Dt
4Dt

satisfies the one-dimensional diffusion equation

u
2u
=D 2
t
x
with the initial condition
u(x, 0) = (x x0 ).

Problem 56. Let A, B be n n matrices. Show that

Z
Z Z
eA+B =
d1 e1 A (1 1 ) +
d1 d2 e1 A Be2 A (1 1 2 )
0
0
0
Z Z Z
+
e1 A Be2 A Be3 A (1 1 2 3 ) +
0

Problem 57. Consider the Hilbert space L2 ([1, 1]). The Legendre polynomials are given by
P0 = 1,

Pn (x) =

where n = 1, 2, . . .. They satisfy

Z +1
Pm (x)Pn (x)dx =
1

1 dn 2
(x 1)n
2n n! dxn

2
m,n ,
2n + 1

Let
(x) =

X
j 0 =0

Find the expansion cofficients dj 0 .

n, m = 0, 1, 2, . . .

dj 0 Pj 0 (x).

Generalized Functions

8.2

79

Supplementary Problems

Problem 1. Show that the 2-dimensional complex -function can be

written as ( C)
Z
Z
1
1
(2)
2

(z) = 2
d exp( z z ) = 2
d2 exp(i( z + z )).
C
C
Problem 2.

Show that
1
(x x0 ) =

1+2

!
cos(kx) cos(kx0 ) .

k=1

Problem 3. Consider the Hilbert space L2 ([0, )). The Laguerre polynomials are defined as
Ln (x) = ex

dn x n
(e x ),
dxn

n = 0, 1, 2, . . .

For the Hilbert space L2 ([0, )) we have the basis

B = { ex/2 Ln (x) : n = 0, 1, 2, . . . }.
Let a R. Show that
(x a) = e

(x+a)/2

Lk (x)Lk (a).

k=0

Problem 4.

Show that

(cos(1 ) cos(2 ))(1 2 ) =

X
`
X

Y`,m (1 , 1 )Y`,m
(2 , 2 ).

`=0 m=`

Problem 5.
that

Let Hk (k = 0, 1, 2, . . .) be the Hermite polynomials. Show

2

e(x +a

(x a) =

Problem 6.

)/2 X
k=0

Hk (x)Hk (a)
2k k!

Show that

 
1
= (z)
z

80 Problems and Solutions

where

Problem 7.

z
2

+i
x
y


.

Let a R. Show that

Z +
Ai(s x)Ai(s a)ds
(x a) =

where Ai is the Airy function.

Generalized Functions

81

Bibliography
Davies E. B.
Linear Operators and their Spectra
Cambridge studies in advanced mathematics, Cambridge University Press,
2007
Dunford E. B. and Schwartz J. T.
Linear Operators. Part 1: General Theory
Interscience, New York (1966)
Golub G. H. and Van Loan C. F.
Matrix Computations, Third Edition,
Johns Hopkins University Press (1996)
Jones D. S.
The Theory of Generalized Functions, Cambridge University Press (1982)
Kato T.
Perturbation Theory of Linear Operators
Springer, New York (1966)
Miller W.
Symmetry Groups and Their Applications
Academic Press, New York (1972)
Reddy B. Daya
Introductory Functional Analysis,
Springer, New York (1990)
Schwartz L.
Theorie des distributions, Hermann, 2 vols. (1966)
Steeb W.-H.
Matrix Calculus and Kronecker Product with Applications and C++ Pro82

Bibliography

83

grams
World Scientific Publishing, Singapore (1997)
Steeb W.-H.
Continuous Symmetries, Lie Algebras, Differential Equations and Computer Algebra
World Scientific Publishing, Singapore (1996)
Steeb W.-H.
Hilbert Spaces, Wavelets, Generalized Functions and Quantum Mechanics
Kluwer Academic Publishers, Dordrecht (1998)
Steeb W.-H.
Problems and Solutions in Theoretical and Mathematical Physics,
Third Edition, Volume I: Introductory Level
World Scientific Publishing, Singapore (2009)
Steeb W.-H.
Problems and Solutions in Theoretical and Mathematical Physics,
Third Edition, Volume II: Advanced Level
World Scientific Publishing, Singapore (2009)
Steeb W.-H., Hardy Y., Hardy A. and Stoop R.
Problems and Solutions in Scientific Computing with C++ and Java Simulations
World Scientific Publishing, Singapore (2004)
Vladimirov V. S.
Equations of Mathematical Physics, Marcel Dekker, New York (1971)
Weidmann J.
Linear Operators in Hilbert Spaces
Springer-Verlag, New York (1980)
Yosida K.
Functional Analysis, Fifth Edition
Springer Verlag (1978)

Index
Admissibility condition, 57
Analysing wavelet, 57
Antilinear operator, 2
Bell basis, 8, 12
Bernstein polynomials, 33, 38
Cauchy sequence, 9
Chebyshev polynomials, 30
Comb, 2
Continuous wavelet transform, 57
Convolution theorem, 53
Dilations, 57
Dirac comb, 74
Fock space, 24, 36
Frame, 2
Fredholm alternative, 3

Laguerre polynomials, 79
Legendre polynomials, 20, 27, 49, 78
Legrendre polynomials, 29
Lie group, 30
Momentum operator, 64
Morlet wavelet, 70
Null space, 3
Paley-Wiener space, 38
Parsevals relation, 43
Peakon, 75
Poisson equation, 77
Position operator, 42
projection theorem, 4
Range, 3
Reproducing kernel, 36
Reproducing kernel Hilbert space, 38
Resolvent, 49

Gaussian function, 54
Gaussian measure, 36
Gram-Schmidt orthogonalization pro- Schauder basis, 35
Shifted Legendre polynomials, 24
cess, 8
Sinc function, 39
Gram-Schmidt technique, 15
Sobolev space, 69
Spectral theorem, 10
Haar scaling function, 57
Spherical harmonics, 66
Hardy space, 6
Symplectic tomogram, 67
Harper operator, 49
Hat function, 52
Translations, 57
Hilbert transform, 59
Triangle function, 35
Inverse wavelet transform, 57

Volterra integral equation, 24

Kernel, 3
Kernel function, 55
84